The citrus industry, one of the most important fruit industries in China, is currently focusing on nondestructive internal quality grading. Near-infrared (NIR) spectroscopy has been widely used to provide accurate analysis of the material components of agricultural products, owing to its advantages of convenience and efficiency. Among different measurement setups, optical fiber probes are frequently used as a key optics accessory for collecting the NIR spectrum. The Monte Carlo (MC) method offers an accurate description of light propagation in the fruit tissue using a multilayer sample model, which provides the theoretical basis for the design of a more effective optical fiber probe in fruit-quality inspection. However, in the MC model, the target sample is usually described as a combination of multiple semi-infinite turbid mediums, which simplifies the structural characteristics of the tissue. Moreover, the structure and geometry of optical fiber probes are also ignored in most cases, which may lead to deviations in the simulation results. This study develops a more detailed description of light propagation in citrus tissues by analyzing the photon transmission near the optical fiber probe as well as the surface boundary of the sphere sample and suggests a probe design for reflectance-based detection.

Based on the characteristics of citrus fruits, a three-layered optical model in the shape of a sphere is established, which includes flavedo, albedo, and vesicle layers with different absorption coefficients, scatter coefficients, refractive indices, and anisotropy factors. A simulation system of citrus-quality detection is developed, comprehensively considering the injection near the source fiber, the photon track near the surface boundary, and the information collected by the detector fiber. The corresponding changes are implemented in the general MC code for the simulation of transmission characteristics in the citrus tissue, including normalized relative diffuse reflectance, average motion pathlength of photons, and the percentage of effective photons in vesicles. Then, MC models with specific parameters of the optical fiber probe, including the numerical aperture and radius of both the source and detector fibers, and the source–detector distance are established for MC simulation.

Based on the simulation results at a wavelength of 800 nm, the guiding principle of the optical fiber probe design suitable for citrus spectrum acquisition is determined. As shown in Fig. 4, the normalized relative diffuse reflectance increases with an increase in the radius of the detector fiber and decreases with an increase of source-detector distance, while other parameters show no significant influence. The average motion pathlength of the received photons increases with increasing source-detector distance (Fig. 5). In this case, the acquired spectrum carries more internal information about the citrus sample. For the percentage of effective photons in vesicles, a large detector fiber radius of 250 μm or more is recommended to achieve a stable result (Fig. 6) as well as a longer source-detector distance. Based on the simulation results, a new structure of “4 in 9 out” coaxial fiber probe is designed to direct light emitted by the light source and to receive the feedback signal (Fig.7). This versatile fiber probe achieves a collecting efficiency of 3.24% and a percentage of effective photons of 0.07% in MC simulation.

Nondestructive inner quality inspection techniques play an important role in the fruit industry. In this study, light propagation through the citrus tissue is simulated by the MC method to determine the relationships between fiber probe geometry parameters and the detected optical signal. A three-layered media model in the shape of sphere is established for further simulation and analysis. The optical fiber probe parameters that are closely related to NIR spectrum measurement are explained and introduced into the general MC model, and corresponding simulations are performed to determine the basic designing principles of the fiber probe in citrus tissues. Based on the simulation results, the radius of the detector fiber probe and the source-detector distance can be optimized to obtain more information of interest, while other parameters including the radius of the source fiber and the numerical aperture show limited impact on the simulation results. To achieve higher collection efficiency, a larger detector radius and smaller source-detector distance are recommended. Moreover, the photons received by the detector fiber are found to carry more information on the inner citrus tissue when the source-detector distance increases. A versatile optical fiber detection structure is designed with a large detector fiber radius and multiple source-detector distances to increase the level of received information of the citrus sample. The MC simulation result of the fiber probe indicates that the photons from the vesicle layer can be efficiently collected and the sensitivity of citrus inner quality detection is further improved, which provides a theoretical reference for designing a new detection accessory of nondestructive quality evaluation by NIR spectroscopy.

Ultrafast laser direct writing processing technology with a single focus has a small processing area and low energy utilization and is not suitable for applications in large-area processing, volume processing, and the single forming of structures. To solve these problems, researchers generally adopt multibeam parallel processing method. The most important aspect of parallel processing technology is to realize three-dimensional (3D) multibeams and form 3D multi-focus with the same energy. Therefore, a method for generating 3D multi-focus is urgently required. In industrial applications, laser focusing requires precise control to obtain high quality results. Moreover, in the actual optical path, setup and manufacturing errors exist in the optical elements and have a significant impact on the final machining results; however, the existing iterative algorithms rarely compensate for such errors. Therefore, a 3D multi-focus control method based on a feedback-weighted 3D-GS algorithm is proposed in this study.

A spatial light modulator (SLM) can adjust optical parameters such as amplitude, phase, and polarization of a laser beam by loading computer-generated holograms. With the help of SLM, the light intensity distribution in the target region can be easily controlled. The key to obtaining uniform 3D multiple beams with SLM is to obtain the corresponding computer-generated holograms, which can then be used to flexibly control the number, position, and focus energy distribution of the outgoing laser beam. In this study, to improve the uniformity of the energy distribution of multiple beams, a feedback method was used to improve the traditional 3D-GS algorithm. Each beam was set using a weight coefficient. The feedback-weighted 3D-GS algorithm collected the energy and position information of multiple foci through a CCD camera in real time and fed the information back to the controlling terminal to dynamically adjust the weight coefficient of each beam. After a couple of iterations, 3D multi-focus with high uniformity were obtained.

The coordinate parameters related to the desired 3D structures in different Z-axis planes are set, and the energy uniformity of the 3D multi-focus is calculated using the feedback-weighted 3D-GS algorithm and traditional 3D-GS algorithm. For the expected 3D structure of “HBUT,” the feedback-weighted 3D-GS algorithm improved the uniformity of 47 diffraction points on four different planes from 47% to more than 96% after 20 iterative feedback calculations. Using the feedback-weighted 3D-GS algorithm, the homogeneity of the 3D multi-focus energy distribution is significantly improved, as shown in Table 1. Another experiment was performed for the 3D spiral structure, and the feedback-weighted 3D-GS algorithm improved the uniformity of 15 diffraction points on 15 different planes from 47% to 94% after five iterative feedback calculations. The spots at the corresponding positions were filled according to the 3D model of the spiral structure (Fig.9). From a comparison of spot energy distribution calculated by different algorithms, it can be found that the presented feedback-weighted 3D-GS method can effectively improve the uniformity of 3D multi foci.

Based on the feedback-weighted 3D-GS algorithm and programmable SLM, this study proposes a method to generate 3D multi-focus with high uniformity to compensate for the fabrication and setup errors of devices in real optical paths. The designed pattern of “HBUT” and helical structures are used to prove the validity of the method. Through the analysis and calculation of feedback parameters in iterative feedback calculations, the uniformity of the 3D multibeam obtained by this method is verified to be 95%. The number of multifocal points in the 3D structure has a greater influence on the stability of the reconstructed 3D multi-focus light field than the number of Z-planes in the 3D structure during the feedback iterative calculation. Additionally, the laser high-uniformity 3D multi-focus optical field reconstruction technique proposed in this study can be used for 3D structure machining.

The advent of the high-speed information age has led to the diversification of user service demand, which the currently available spectrum cannot meet. Hence, exploiting higher-frequency millimeter wave signals has become necessary. However, the transmission performance of high-frequency millimeter wave signals is unstable, and their transmission distance is significantly shorter. Radio-over-fiber (ROF) technology using optical fibers to transmit wireless signals has low attenuation loss, large capacity, and long transmission distance. A combination of wavelength division multiplexing (WDM) technology with ROF, WDM-ROF, can handle the diversification of services and load different data in multiple channels. In recent years, it has become an area of intense research interest.

The current WDM-ROF systems have some issues, such as the high cost of the laser array, poor scalability and reconfigurability, generation of single and non-adjustable millimeter wave signals, same service transmission for each base station, and being applicable only to a small number of scenarios. To overcome these shortcomings, we propose a multi-service layered WDM-ROF system with an optional frequency millimeter wave based on an optical frequency comb.

The structural diagram of the proposed multi-service layered WDM-ROF system with an optional frequency millimeter wave based on an optical frequency comb is shown in Fig. 1. The system consists of CS (central station), RN (remote node), SCS (subcentral station), and BSG (base station group). In the CS, the generation of a flat optical frequency comb, loading of data, and uplinking of a reserved optical carrier are accomplished, which reduces the cost of the laser array and improves its scalability and reconfigurability. In the RN, the required optical frequency is filtered out by a WDM filter. In the SCS, the optical switch matrix controlling the combined output of different optical frequencies can realize the selection of millimeter wave frequencies. In the BS, four millimeter wave signals with different frequencies and two service information are generated by the PD and transmitted through the electric amplifier (EA) antenna, reducing the cost due to identical transmission in each base station and making the system applicable to more situations. Simultaneously, the uplink signal is loaded to realize uplink transmission. The proposed multi-service layered WDM?ROF system with optional frequency millimeter waves based on an optical frequency comb was simulated and verified using OptiSystem simulation software. In addition, its transmission sensitivity and power cost were studied.

The proposed optical frequency comb generation scheme meets the requirements of a 21-line optical comb with a frequency interval of 20 GHz, and flatness of 0.96 dB is obtained by simulation (Fig. 2). In the downlink, two 10 Gbit/s downlink data are upconverted to four-millimeter wave signals of 35, 45, 65, and 95 GHz using the MEMS optical switch matrix in one of the control states and PD, and the bit error rate (BER) curves of four millimeter wave signals after B-T-B, 20 and 40 km SMF transmissions are shown (Fig. 6). The figure shows that the BER increases gradually with a reduction in the received optical power, and the relationship between the receiving sensitivity and BER is approximately linear. The receiving sensitivity and power cost of four millimeter wave signals in the downlink after 20 and 40 km SMF transmissions are listed in Table 1. The table shows that the receiving sensitivity gradually decreases, and the power cost gradually increases with an increase in the transmission distance. The receiving sensitivity slightly degrades with an increase in the millimeter wave signal frequency, the power cost slightly increases (within 1 dB) after 20 and 40 km SMF transmissions. The BER curves of uplink data updata1 and updata2 after B-T-B, 20 and 40 km SMF transmissions are shown in Fig. 8. The figure shows that the BER curves of uplink data updata1 and updata2 are similar to those of the downlink, the BER gradually increases with a decrease in the received optical power, and the receiving sensitivity is linear with the BER. The receiving sensitivity and power cost of updata1 and updata2 after 20 and 40 km SMF transmissions are given in Table 2. The table shows that the receiving sensitivity gradually decreases, and the power cost gradually increases with an increase in the transmission distance. The receiving sensitivities of updata1 and updata2 are similar, and the power costs after 40 km SMF transmission are 0.846 and 0.837 dB, respectively. The transmission performance of the uplink is good and better than that of the downlink.

This study proposes a multi-service layered wavelength-division multiplexing radio-over-fiber (WDM-ROF) system in which an optional frequency millimeter wave is generated based on an optical frequency comb (OFC). The transmission sensitivity and power cost characteristics of the WDM-ROF system were studied. The simulation results show that the average power loss of the downlink and uplink is less than 1 dB after 40 km of SMF transmission, while the eye diagrams of the downlink and uplink remain open after 40 km of SMF transmission; the transmission performance of the uplink is better than that of the downlink. By increasing the number of optical frequency comb lines, adjusting the frequency interval of the optical frequency comb, increasing the optical switch matrix structure, and configuring multiple groups of base stations, the system can realize BSG expansion, frequency selection, reconstruction of millimeter wave signals, and multiple service transmissions, providing an effective way for next-generation broadband optical wireless access networks.

Currently, the information capacity of communication systems based on a single-mode fiber (SMF) is approaching its physical limits. To solve this problem, spatial division multiplexing based on orbital angular momentum (OAM) has been intensively investigated. Theoretically, owing to their orthogonality characteristics, OAM modes can help realize infinite multiplexed channels. Hence, the capacity and spectrum efficiency of existing optical fiber communication can be enhanced by using the orthogonality and infinity properties of the OAM modes. The vortex mode amplifier is one of the key devices in optical communication systems and is essential for improving the performance of future vortex mode multiplexing systems applicable in long-distance and high-capacity optical fiber communication. Currently, the vortex mode amplifier heavily relies on the ring-core erbium-doped fiber (RC-EDF), which is not only conducive to the stable transmission of the vortex beam but also improves the conversion efficiency of the pump beam. The modal gain is an important index of ring-core erbium-doped vortex fiber amplifiers. Therefore, the design and fabrication of an RC-EDF with a high modal gain are essential to satisfy the growing demand for data traffic.

The modal gain mainly depends on the intensity distribution of the pump and signal modes as well as the doping distribution of erbium ions. The ring core structure is used to optimize the field distribution of the pump modes, thereby improving the conversion efficiency of the pump beam. Owing to the limitations of the doping technology, the doping distribution of erbium ions in actual fabricated fibers differs considerably from that of the theoretical design. Therefore, in this study, we simplify the Er doping factor in the theoretical design to ensure that the doping distribution of the erbium ions is uniform. First, the effects of the doping region and width of the ring core on the gains of the first- and second-order vortex modes are analyzed, and subsequently, these effects are optimized to realize a high-gain ring-core erbium-doped vortex fiber amplifier. According to the relative refractive index profile of the RC-EDF, the modal intensity profiles of the vortex modes (|l|=1-2) and the pump fundamental mode (PFM) are numerically simulated using the finite-element method. The effects of the RC-EDF length and signal wavelength on the amplifier gain characteristics are studied using numerical simulation based on the rate and light propagation equations, which guides the optimized fiber parameters in RC-EDF fabrication. Furthermore, an experimental setup is built to characterize the amplification performance of the proposed RC-EDF.

The proposed RC-EDF has cladding and inner-ring-core radii of 62.5 μm and 1.6 μm, respectively. We ensure a large refractive-index difference of 0.012 between the ring core and cladding, which is sufficient to achieve mode splitting. The proposed RC-EDF supports high-order vortex modes (|l|=1-2). Modal gains (|l|=1-2) are optimal when the thickness of the ring core and the width of the doping region are 5 μm and 6 μm, respectively. The simulation results show that the gains of the first- and second-order vortex modes can reach 35.4 dB in the C band [Fig. 2(b)]. The fabricated RC-EDF has a cladding radius of 62.5 μm, an inner ring-core radius of 1.7 μm, and a ring core thickness of 5 μm. Moreover. the refractive-index difference between the ring core and cladding is 0.0127. An experimental setup is built to characterize the amplification performance of the first- and second-order vortex modes in the proposed RC-EDF. Furthermore, an infrared camera is used to detect the output beam profile of the RC-EDF. The results show that the clean vortex modes (|l|=1-2) are stably excited and transmitted in the RC-EDF. The maximum gain (|l|=1-2) is 32.6 dB at 1530 nm [Fig. 7(b)]. The maximum output power values of the first- and second-order vortex modes are 14 dBm and 15 dBm, respectively (Fig. 8). The fabricated RC-EDF has a small overlap area between the pump and signal modes, and the doping distribution of the erbium is uneven, which causes the gains to be different in the theoretical calculations and experimental measurements.

In this study, a ring-core erbium-doped fiber that supports the first- and second-order vortex modes is designed, and modal gains (|l|=1-2) are improved by optimizing the thickness of the ring core and the width of the erbium-doping region. Based on these optimized parameters, we perform a theoretical simulation and find that the vortex mode gains are higher than 35.4 dB in the entire C band. The performance of the fabricated RC-EDF is experimentally characterized. The gains of the vortex modes (|l|=1-2) are higher than 32.6 dB at 1530 nm wavelength. The proposed high-gain vortex mode amplifier based on the ring-core erbium-doped fiber is expected to be widely used in long-distance and large-capacity spatial division multiplexing fiber communications.

A large-capacity, high-security quantum noise random cryption (QNRC) system requires high-speed, high-resolution pseudo-multi-ary signal waveforms. However, the generation of pseudo-multi-ary signal waveforms requires a high-speed, high-resolution digital-analog converter/analog-to-digital converter (DAC/ADC). Therefore, a high-speed, high-resolution DAC/ADC plays a crucial role in the performance of QNRC systems. However, owing to the performance limitations of the current high-speed, high-resolution DAC/ADC, the performance of the QNRC system is limited. Our suggested approach is motivated by the need to utilize a new method based on low speed and low resolution to avoid the usage of high-speed, high-resolution DAC in the QNRC system, hence eliminating the performance constraint of QNRC systems owing to the DAC bottleneck. Therefore, the potential of QNRC systems can be fully evaluated while also lowering the system cost.

In this study, a flexible multi-ary PSK-QNRC system based on a low-speed, low-resolution DAC combined with cascaded phase modulators is proposed. The proposed system is optical domain decryption based on coherent detection, which is simple and easy to implement. At the transmitter of the system, the encryption mapping of a 12 bit running sub-key with an n bit plaintext signal is performed on a bit-by-bit basis (if n<4, we only need to place (4-n) digits with “0” in front of them), where the first group of four-bit-DAC is used to modulate the first phase modulator, the second group of four-bit-DAC is used to modulate the second phase modulator, and so on. Four phase modulators are used in our experimental system, which are connected by 100 ps delay lines. The cascaded modulated signal output is attenuated to a mesoscopic coherent state, which has a power of -20 dBm, by an optical attenuator and then sent to the transmission link of the PSK-QNRC system. At the receiver end, the optical signal carrying the decryption information of the running sub-key is used as the reference light (LO, local oscillator) for coherent demodulation, and the ciphertext signal transmitted through a span of the optical fiber with dispersion compensation is used as the signal light (SIG) for coherent demodulation. These two optical signals are then simultaneously sent to the coherent receiver for decryption after time-delay matching. The output signals from the coherent receiver include an in-phase branch (I) signal and a quadrature branch (Q) signal. The I and Q branch signals are then sent to a real-time oscilloscope, where signal phase estimation is performed. Finally, bit error rate estimation is performed based on the results of the signal phase estimation. Through theoretical and experimental analyses of the PSK-QNRC encryption system, the feasibility of the proposed scheme for a large-capacity, long-distance quantum-noise random-encryption transmission system is verified.

To evaluate the proposed scheme, we established an experimental optical PSK-QNRC system (Fig. 3) and a corresponding computer simulation system based on VPI9.1 software (Fig. 4). All the parameter configurations are listed in Table 1. Taking the transmission of the binary plaintext signal (1+12) as an example, the simulation results are shown in Fig. 5. After the ciphertext signal is transmitted and decrypted over a long distance, the signal is restored to a binary signal, and a clear eye diagram is obtained. A legitimate receiver can accurately obtain the plaintext information, whereas an illegal eavesdropper cannot obtain the transmitted signals from the encrypted signal. We discussed and analyzed the performance of the proposed scheme and compared it with a traditional scheme using high-speed and high-resolution ADC/DAC under binary plaintext. Additionally, we analyzed the security performance of the proposed system under binary and multi-ary decryption. The system transmission performance of the proposed scheme is analyzed under a multi-ary decryption setting and compared with the traditional scheme. Figure 6 shows the power penalty comparison curve between this scheme and the 16 bit high-speed, high-resolution DAC scheme under a binary plaintext setting. The results show that the proposed system can retain the performance of the traditional scheme while avoiding the performance limits of the high-speed QNRC transmission system imposed by the DAC resolution limits and can significantly reduce the system cost. Figure 7 shows the evaluation results of the system's security performance. It can be clearly observed that the NMS (number of quantum state) values under the binary decryption and multi-ary decryption schemes (where 4-ary, 6-ary, and 8-ary are considered) are 2.33, 4.66, 9.33, and 18.66, respectively, which can meet the security performance requirements of the system in both binary and multi-ary decryption settings. Figure 8 shows the time-domain waveforms, eye diagrams, and constellation diagrams of the decrypted signals when the plaintext is quaternary. It can be observed that after decryption, both the binary and multi-ary signals are successfully recovered, and the eye diagram is clearly visible. A legitimate receiver can obtain the corresponding original plaintext signal under each condition after signal decryption; however, an illegal receiver cannot obtain the original plaintext information from the encrypted signals. Figure 9 depicts the system bit error rate curve of the proposed scheme with a 10 km signal transmission under multi-ary settings. The results confirm the feasibility of the proposed scheme for multi-ary decryption system applications.

To solve the performance limitation of high-speed and high-resolution DAC on the performance of the QNRC system, this study proposed a scheme for designing a high-bit QNRC transmission system based on a low-speed, low-resolution DAC combined with cascaded phase modulators. The proposed scheme not only overcomes the transmission performance constraint imposed by the DAC bottleneck in cascaded PSK-QNRC systems but also significantly reduces the system cost. Moreover, the proposed scheme can be adapted for multi-ary transmission applications.

First, the proposed scheme was discussed, analyzed, and compared with the traditional 16 bit high-speed scheme with a high-resolution DAC in terms of the power penalty. Subsequently, the feasibility of the proposed scheme was verified. Second, this study applied this scheme to a multi-ary transmission setting and conducted a comparative analysis of the system's performance under various multi-ary transmission conditions. The feasibility of the future QNRC system adapting to a multi-ary transmission system was verified. Third, the domestication of DAC with a 4 bit resolution and 30 Gb/s transmission rate used in the proposed system has already been realized, which can bypass the dependence on the imported high-speed, high-resolution DACs, thus providing a feasible solution for realizing domestic QNRC systems with low system costs. Notably, the system proposed in this study is scalable because the maximum number of ciphertext states in the transmission is not limited to 16 bit. When necessary, it can be upgraded by cascading more low-speed DACs at the end of the transceiver or a few higher-resolution DACs to increase the transmitted ciphertext state of the QNRC system.

A frequency converter, which is an important part of the receiver and transmitter in communication systems, is widely used in broadband wireless communication, radar, satellite communication, etc. Because signals are processed in the optical domain, photonics-based converters possess several advantages, such as large bandwidth, low loss, and anti-electromagnetic interference, thus providing a new solution for modern communication systems. The frequency conversion functions of frequency converters include up-conversion, down-conversion, and in-phase/quadrature (I/Q) up-conversion. Most proposed schemes may implement one of these frequency conversion functions, which present limited application significance. Therefore, researchers have proposed the generation of multiple frequency-converted signals in the same structure. Nevertheless, optical bandpass filters are widely used in these schemes to filter out unwanted optical sidebands, which severely limits the frequency coverage of the frequency converter. Meanwhile, complicated operations (including changing the input signal and adjusting the direct current (DC) bias voltage or 90° electrical phase shift of the modulator) must also be implemented in these approaches. We propose a reconfigurable dual-output microwave photonic frequency converter with high frequency tunability. Such a converter may perform multiple frequency conversion functions with dual outputs, and it is expected to meet the demands of future multifunctional and wide-bandwidth communication systems. Up- and down-converted signals, up-converted upper and lower sideband signals, and vector signals can be generated by changing the input signal in the reconfigurable structure. Moreover, the frequency-converted signals may be output separately in two channels, and the converter is filterless.

In our proposed scheme, an optical carrier generated by a laser diode is transmitted to a polarization division multiplexing dual-parallel Mach-Zehnder modulator (PDM-DPMZM). Moreover, in addition to a local oscillator (LO) signal, an intermediate frequency (IF), a radio frequency (RF), or an I/Q baseband signal generated by an arbitrary waveform generator is loaded into the PDM-DPMZM. Then, the output signal of the PDM-DPMZM is transmitted to a polarization controller (PC). The PC has a polarization rotation angle of 45° and a phase difference of 90°. The positive/negative sideband of the LO signal is eliminated by the PC in two orthogonal polarization directions. A polarization beam splitter is employed to implement polarization separation, which splits the input signal into two parts for photoelectric conversion. Then, two frequency-converted signals are separately generated by the photodetectors in two independent channels. Our reconfigurable scheme can generate up- and down-converted signals, up-converted upper and lower sideband signals, or vector signals when the input signal is changed.

The input RF signal has a uniformly spaced frequency ranging from 9 to 29 GHz, whereas the LO signal has a fixed frequency of 30 GHz. After the beat frequency, up- and down-converted signals with equal frequency intervals are obtained simultaneously and independently. The generated signals present an electrical spurious suppression ratio (ESSR) of 31 dB and a frequency range of 1-59 GHz (Fig. 4). Up-converted upper and lower sideband signals can be generated by changing the RF signal to an IF signal. The results show a highly flat power response for the generated signals (Fig. 5). To verify the feasibility of generating a vector signal with a high frequency, an I/Q baseband signal with a rate of 400 Msym/s was applied to implement I/Q up-conversion. Thus, 64-quadrature amplitude modulation (QAM) signals centered at 30 GHz can be obtained (Fig. 6). This implies that our proposed reconfigurable scheme can generate multiple frequency-converted signals by switching the input signals. Moreover, frequency-converted signals can be obtained simultaneously in two independent channels. These results are consistent with those of the theoretical analysis. The error vector magnitude (EVM) of the 64QAM signal was evaluated (Fig. 7). The EVM value is observed to fluctuate from 2.67% to 3.26% when the frequency ranges from 5 GHz to 40 GHz. This indicates the good frequency tunability of the 64QAM signals. The transmission performance was also evaluated, which considers the situations of back-to-back and 30 km single-mode fiber transmission (Fig. 8). The measured EVM values of the upper and lower sideband signals of the up-conversion frequency are both less than 4.5% with minimal fluctuations. This indicates good transmission performance and suitability for long-distance optical fiber transmission.

In this paper, we propose a reconfigurable dual-output microwave photonic frequency converter that is capable of realizing multiple frequency conversion functions, such as up-conversion, down-conversion, and I/Q up-conversion. Up- and down-converted signals, up-converted upper and lower sideband signals, and vector signals can be generated by changing the input signal. Frequency-converted signals can be generated simultaneously in two independent channels with an ESSR greater than 30 dB. Such a filterless scheme presents good frequency tunability as well as a large frequency range of 1-59 GHz. 64QAM signals centered at 5-40 GHz with EVM values less than 3.5% can also be obtained when the I/Q baseband signal is applied. Moreover, our scheme may eliminate the periodic power fading effect caused by fiber dispersion. This indicates its suitability for long-distance fiber transmission. Its spurious-free dynamic range (SFDR) is as high as 107.1 dB·Hz^2/3. The performance interference caused by non-ideal factors was also evaluated and analyzed, and the results demonstrate the practicability and feasibility of our proposed scheme.

With the rapid development of global industry, air pollution is now a major problem. Ammonia (NH₃) is a common toxic and harmful gas that is used in industrial production. Excessive intake of NH₃ by the human body can lead to lung swelling and even death. Currently, NH₃ detection is mainly conducted using electrochemical and optical methods. However, traditional electrochemical detection methods cannot realize real-time online monitoring. Optical detection can be classified into three types: absorption-spectrum, evanescent-wave, and refractive index variation. The absorption-spectrum type is easily affected by other gases in the same absorption band, whereas the evanescent-wave type has high requirements for achieving a precise tapering process. The refractive index variation type is based on variations in the refractive index of the sensitive film derived from NH₃. In recent years, sensing technology that combines optical fiber sensors and functional films for realizing specific gas detection has become a research hotspot. Because zinc oxide (ZnO) has strong adsorption characteristics for NH₃, the refractive index of ZnO varies. A no-core fiber is an optical waveguide with a unique structure, and its mode field can directly perceive changes in the external environment. Therefore, a new fiber gas sensor that combines a no-core fiber and ZnO film is studied in depth, which significantly promotes the rapid and accurate measurement of NH₃ in atmospheric pollutants.

Based on mode transmission theory, the mode characteristics of a ZnO-coated no-core fiber and the sensing characteristics of a singlemode-no-core-singlemode (SNS) structure were analyzed using MATLAB software. First, the mode field distribution in the no-core fiber coated with ZnO was studied, and variations of the refractive index of the ZnO film with mode excitation coefficient and self-image length were discussed. Second, the relationship between the refractive index of the ZnO film and resonant wavelength of the interference spectrum was established. The NH₃ sensing system mainly includes a light source, glass chamber, and spectrometer. Finally, in an experiment, different volume fractions of NH₃ were passed into the glass chamber to ensure the sensor made full contact with NH₃. When ZnO absorbs NH₃, its refractive index changes, and thus the resonant wavelength shifts. In our study, NH₃ volume fraction was detected by establishing the relationship between the NH₃ volume fraction and the shift in resonant wavelength. The detection limit of the sensor was calculated using the Hubaux-Vos method, and the response time and recovery time were tested. In addition, the ambient temperature was controlled using a temperature-control box to study the effect of temperature on NH₃ sensing.

Changes in the refractive index of a ZnO film affect to some degree the mode excitation coefficient and self-image length in a no-core fiber coated with a ZnO film layer. The sensitivity of the sensor increases with increasing film thickness. In our study, the sensitivities of the SNS sensor with thicknesses of 60 nm and 130 nm are 11.8 and 28.6 nm/RIU, respectively (Fig.6). The prepared ZnO film was characterized using scanning electron microscopy (SEM), and the results show good compactness and uniformity (Fig.7). At 16.5 ℃, the resonant wavelength blue-shifts with an increase in NH₃ volume fraction. The sensitivities of the SNS sensor at ZnO film thicknesses of 60 and 130 nm are 17.96×10⁶ and 17.86×10⁶ pm, respectively, which is mainly caused by the effects of ZnO on NH₃ adsorption saturation (Fig.9). The average sensitivity of the SNS sensor under NH₃ detection coated with a ZnO film at a thickness of 60 nm is 16.87×10⁶ pm, and the detection limit is 6.6×10^-6 (Fig.10). The effect of time on the detection limit is 0.026×10^-6 d^-1, and the response and recovery time are 70 and 90 s, respectively (Fig.11). The sensitivities of the sensor under NH₃ detection are 17.96×10⁶, 15.38×10⁶, 14.07×10⁶, 9.62×10⁶, and 7.57×10⁶ pm at 16.5, 25, 30, 45.2, and 56 ℃, respectively (Fig.12). The effect of temperature on the detection limit is 0.0237×10^-6 ℃^-1. The detection sensitivity of NH₃ decreases with an increase in temperature, which derives from the fact that the increase in the surface potential barrier of the ZnO film hinders the flow of electrons. In practical sensing applications, measuring the temperature of the environment is first required, and then the volume fraction of NH₃ can be calculated based on the data interpolation.

A high-sensitivity NH₃ sensor based on a ZnO-coated SNS structure was proposed in this study. SNS sensors with different ZnO film thicknesses were simulated theoretically. The results show that the mode field distribution trend of the no-core fiber coated with ZnO is consistent. When the refractive index of the ZnO film changes from 1.929 to 1.889, it has little effect on the excitation coefficient and self-image length. In general, the sensitivity of an SNS sensor increases with an increase in film thickness. In this study, an SNS sensor with ZnO film thicknesses of 60 and 130 nm was prepared by atomic layer deposition (ALD), with results showing sensitivities of 11.8 and 28.6 nm/RIU, respectively. The results show that the sensitivities to NH₃ are essentially the same under the two film thicknesses (17.96×10⁶ and 17.86×10⁶ pm, respectively), which mainly derives from the effects of ZnO on NH₃ adsorption saturation. The average sensitivity, detection limit, and response and recovery time are 16.87×10⁶ pm, 6.6×10^-6, and 70 and 90 s, respectively. Finally, an in-depth analysis of the temperature characteristics of the sensitivity of the sensor demonstrated that NH₃ detection sensitivity decreases with an increase in temperature. At 56 ℃, the sensitivity decreases to 7.57×10⁶ pm. The effects of temperature and time on the detection limit of NH₃ are 0.0237×10^-6 ℃^-1 and 0.026×10^-6 d^-1, respectively.

With the rapid development of mobile communication technology in modern society, the demand for location services in complex indoor environments, such as large factories, shopping malls, and office buildings, has been growing rapidly. The current visible light positioning technology uses various sensors and hybrid complex algorithms to achieve positioning, which is difficult to operate and vulnerable to interference, resulting in unstable positioning accuracy of the system. The advantages of visible light communication include both lighting and communication, as well as stability and reliability. On this basis, to improve the accuracy and stability of visible light indoor positioning, a bio-inspired network integrating migration feature learning is proposed to achieve stable and high-precision indoor positioning in visible light imaging.

In this study, a visible-light indoor location method based on an image is proposed. The acquired image is first denoised to eliminate noise interference which has a significant impact on the extraction of the image depth features. Inaccurate feature extraction leads to poor positioning accuracy. An improved threshold denoising method is used to address the issue of signal loss caused by the oscillation of the threshold function. The adjustment function ensures good continuity of the signal and retains the original features of the image to the maximum extent based on image denoising. Second, the ResNet network is used to extract image depth features and establish a fingerprint database. The image depth features exhibit translation and rotation invariance. However, the ResNet network has deeper network layers than the traditional neural networks. Thus, residual learning is added to avoid a decrease in accuracy resulting from the increase in network layers. Finally, the BAS algorithm is used to optimize the connection weight matrix between the layers of the RBF neural network, improve the training speed and stability of the network, and determine the optimal weight between the layers of the network through back propagation for enhanced positioning accuracy.

In this study, we first build a positioning experimental platform (Fig. 4) consisting of a light environment that can simulate real indoor scenes to verify the applicability and effectiveness of the algorithm. The coordinate plate at the bottom of the experimental box is divided into several areas of equal size at an interval of 5 cm, and four LED light sources with the same size and power are installed on top of the experimental box to collect visible light images and extract depth features. Pictures are collected at three different heights by lifting and lowering the coordinate plate to establish a depth feature database for the collected pictures. The measured data are input into the neural network for training. The RBF neural network achieves 26983 target error iterations, the BAS optimized RBF neural network achieves 47352 iterations (Fig. 5), and the training speed is increased by approximately 40%. We randomly select 30 different coordinate points in the experimental box and collect the corresponding images without denoising for the positioning test. The average positioning error without denoising is 5.02 cm (Fig. 6), whereas, with denoising applied to the images collected at the same 30 points, the average positioning error is 4.26 cm (Fig. 7). The experiment shows that image denoising can effectively improve positioning accuracy. When compared to the RBF and back-propagation (BP)network algorithms, the BAS-RBF neural network algorithm provides significant improvements (Fig. 8) . Compared with the BP network algorithm, the confidence probability of a fixed error of less than 2 cm, 4 cm, and 6 cm increase by 9%, 11%, and 10%, respectively. The experimental results show that the performance of the RBF neural network optimized by BAS is better than those of the RBF and BP neural networks (Table 3). The average positioning error of the algorithm is 4.26 cm, which is 10.5% higher than that of the RBF neural network and 16.9% higher than that of the BP neural network.

This study proposes a visible light positioning technology for visual imaging that requires only images from indoor locations. Subsequently, the migration feature learning is used to extract the depth features of the denoised images to establish a database, which is brought into a neural network fused with biological algorithms for learning and training, with the goal of building a neural network training and testing model. Compared with the RBF and BP networks, this model can improve the positioning accuracy and training speed. In the actual measurement and positioning stage, 0.8 m×0.8 m×0.8 m physical model, the average positioning error of the prediction result is 4.26 cm; the probability of the prediction point error of less than 4 cm is 63.4% and the probability of the prediction point error of less than 6 cm is 78%. The positioning result is stable and reliable, providing a new feasible scheme for visible-light indoor positioning technology.

High-speed modulated semiconductor lasers are important light sources for high-capacity optical communication systems. Compared with externally modulated lasers, such as electro-absorption modulated distributed feedback (DFB) lasers, directly modulated DFB lasers have several advantages, including a simple structure, low cost, and low power consumption. A higher speed of data transmission of an optical communication system can be obtained using DFB lasers with a higher direct modulation bandwidth in the following ways. First, high-gain active materials such as InGaAlAs/InP multi-quantum wells (MQWs) can be used for the fabrication of lasers. Subsequently, a short laser cavity length can be used to realize a short photon lifetime in the cavity. Because of their important applications, high-speed directly-modulated InGaAlAs/InP MQW DFB lasers have been widely studied. However, to obtain a high modulation bandwidth, the length of the active region for most reported lasers must be less than 150 μm. A small active length leads to a high facet loss, and thus, a low optical power output. In addition, a small length results in high resistance, which leads to a strong self-heating effect. In this paper, we report high-speed directly-modulated DFB lasers integrated with a distributed Bragg reflector (DBR) section working at 1.3-μm wavelength. For the cavity length of 200 μm, the 3-dB small signal direct modulation bandwidth of the laser is larger than 29 GHz.

This device is fabricated via two-step lower pressure metal organic chemical vapor deposition (MOCVD) growth. In the first step, the active layer, a multi-quantum well structure comprising nine 1.2% compressively strained InGaAlAs wells and ten 0.2% tensile-strained InGaAlAs barriers, is grown. The thicknesses of each well and barrier are 4 nm and 10 nm, respectively. A 50-nm-thick InGaAlAs graded index layer and a 50-nm-thick InAlAs laser are grown on both sides of the MQW layer. A 60-nm-thick InGaAsP layer is grown on the upper InAlAs layer for grating fabrication. After a uniform grating is fabricated using electron beam lithography and dry etching, an InP cladding layer and an InGaAs contact layer are grown in the second growth step. Figure 2 shows a schematic of the cross-section structure and an optical graph of the fabricated laser. The laser has a 1.7-μm-wide ridge waveguide structure and consists of a 200-μm-long DFB section and 130-μm-long DBR section. The gratings of the two sections have the same period and etching depth. To obtain a high single-mode yield, a λ/4 phase-shift structure is placed in the middle of the DFB section. The light emitted from the DFB section is reflected back by the DBR section, which helps increase the optical power. Moreover, the feedback from the DBR section can further increase the yield of single-mode emission. As shown in Fig. 2(a), the two sections of the device have the same InGaAlAs MQWs, which greatly simplifies device fabrication.

The threshold current of the laser at 20 ℃ is 12 mA. The optical power is 18 mW at the injection current of 100 mA [Fig. 3(a)]. The emission wavelength of the laser is approximately 1320 nm. The side-mode suppression ratio of the optical spectra is larger than 50 dB when the DFB current increases from 30 mA to 90 mA [Fig. 3(b)]. At 20 ℃, the 3-dB small signal direct modulation bandwidth of the laser is 20 GHz at the injection current of 60 mA and increases to 29 GHz when the current increases to 90 mA [Fig. 4(a)]. By fitting the experimental modulation response curve, the intrinsic modulation bandwidth is found to be 35.4 GHz at the current of 94 mA, indicating that the modulation bandwidth of the laser can be further enhanced by optimizing the laser design and fabrication process. The frequency of the relaxation oscillations, which is also obtained by fitting the experimental modulation response data, increases with the injection current at the rate of 0.25 GHz/mA. 25-Gbit/s nonreturn to zero (NRZ) data transmissions using the laser are conducted. Data patterns having a 2¹⁵-1 length are generated by a commercial pulse pattern generator. Clear open eye views can be obtained after 10 km transmission (Fig. 5). The bit error rate (BER) performance of the transmission is also analyzed. The 10-km transmission power penalty of obtaining BER of 10^-10 is less than 1 dB at both 20 ℃ and 80 ℃ (Fig. 6).

A high-speed directly-modulated DFB laser integrated with a DBR section working at 1.3-μm wavelength is fabricated using InGaAlAs/InP multi-quantum wells as the active material. For the cavity length of 200 μm, the 3-dB small signal direct modulation bandwidth of the laser is larger than 29 GHz. Under 25-Gbit/s NRZ data direct modulation, the power penalty to obtain the BER of 10^-10 after single-mode fiber transmission of 10 km is less than 1 dB at both 20 ℃ and 80 ℃. A longer active region length of the device is beneficial for improving the output slope efficiency and reducing the adverse effects of current heating. The fabricated device is a promising light source for short-reach, high-capacity optical communication systems.

The Brillouin random fiber laser (BRFL) is a new type based on stimulated Brillouin scattering (SBS) and a randomly distributed feedback resonator. Because SBS enables low-intensity noise, low-phase noise, and narrow-linewidth lasing light, it has significant advantages in random fiber laser construction. However, research on BRFL has been limited to nonpolarization parameters, such as lasing power, intensity phase noise, and line width, rarely to polarization properties. In this study, a BRFL with a polarization-maintaining fiber (PMF) line-cavity (PMF-BRFL), in which lasering light with polarization clamped at either one of the two orthogonal principal axes of the PMF, is proposed and demonstrated based on the nonlinear axial polarization pulling effect of SBS in PMFs. A theoretical model of the PMF-BRFL is established, and the polarization properties of the lasing light related to the pump light and system parameters are analyzed, discussed, and compared with the experimental results, which were in good agreement with each other.

First, based on the simplified polarization vector-propagation equations of SBS in PMFs, which theoretically indicate the axial polarization-pulling behavior of SBS in PMFs, we derived the analytical expression of the SBS gain in the PMFs, which presents the SBS gain expression to the input SOPs of the pump and signal light and the input pump power. Second, we derived lasing-pump power thresholds for the two polarization modes. We then analyzed the working conditions of these polarization regions. Furthermore, the width of the depolarization range W was analyzed, and its relationship with the pump power and cavity length was discussed.

The PMF-BRFL used a tunable laser to output the pump light with a center wavelength of 1553.73 nm, a polarization state generator (PSG) to generate 100-input SOPs of pump light with a relatively uniform distribution on the Poincaré sphere, an erbium-doped fiber amplifier (EDFA) to vary the input pump power, and a polarization controller to adjust the relative position of p^in to the principal axis of the PMF before the pump light entering a 3 km PMF fiber line-cavity via a normal single-mode fiber circulator (Cir). In a random cavity, the PMF acts as an SBS gain medium and provides the first and second Rayleigh scattering (RS) in opposite directions for random reflection. The SBS Stokes the light generated in the PMF emitted through Cir. At the PMF-BRFL output, the polarization behaviors, including Stokes parameters and degree of polarization, lasing power, optical spectrum, and linewidth of the lasing light, were measured using a polarization analyzer (PSA), an optical spectrum analyzer (OSA), and an electronic spectrum analyzer (ESA).

When the cavity length increases from 1 to 5 km, the lasing threshold of the pump power at zero depolarization decreases from 75 to 29 mW, and the dynamics of the pump power for W<0.1 requirement decreases from 7 to 2.5 mW. For more extended cavities (5-11 km), I_p0th,W=0 decreases slightly, only from 29 to 21 mW, and the dynamics of the pump power for W<0.1 requirement remains almost unchanged around 2.5 mW (Fig. 2). Therefore, the cavity length of the established PMF-BRFL system can achieve an increased pumping efficiency by selecting a PMF of 3-5 km.

For different settings of power emitted from the EDFA, I_{p0_edfa}=40, 50, 55, 80, and 120 mW. As the input SOP of the pump light is scanned in the generated p^in#100 pattern, almost all the SOPs of the lasing light clamp at either of the ±β^l positions and show a definite relationship of sign(p^in?β^l)β^l, except for the cases for W of p^in?β^l≈0. For lasing lights clamped at ±β^l polarization, the system works in the full polarization regions, with measured DOP=1. However, for lasing lights with DOP<1, the system operates in the depolarization region, and both ±β^l modes oscillate simultaneously with different quantities, resulting in different DOPs observed inside the Poincaré sphere (Fig. 5).

W attains the lowest value at I_{p0_edfa}=55 mW, indicating I_p0th,W=0 of the PMF-BRFL with 3 km PMF is near 55 mW. At I_{p0_edfa}=55 mW, depolarization interval W is the narrowest, and the polarization clamping range of the system is the broadest. At I_{p0_edfa}<55 mW, a spontaneous region exists around p^in?β^l≈0, where PMF-BRFL emits weak and depolarized light. At I_{p0_edfa}>55 mW, the depolarization region also increases, and W increases with increasing pump power. The relationship between W and I_p0,W=0 of the PMF-BRFL was measured, and assuming that the Cir insert loss is 1.87 dB, the measured curve is consistent with the theoretical curve for a cavity length of 3 km (Fig. 6).

Moreover, the variation in the lasing power with p^in?β^l was determined, and the spectra of the lasing light were measured for different p^in?β^l values. The lasing power decreases with decreasing p^in?β^l, with a power difference of ~15 dBm between p^in?β^l=1 and 0 at the working point of I_{p0_edfa}=55 mW. Finally, the linewidths of the lasing lights in the two polarization modes are measured for p^in?β^l=1 and 0, and all lights have a narrow linewidth of ~0.75 kHz (Fig. 8).

In this study, a BRFL with bistable orthogonal polarization was proposed and achieved based on the axial polarization pulling effect of the SBS effect in PMFs. First, the polarization mode working regions of the PMF-BRFL system and the corresponding working conditions were analyzed and discussed. In addition, a PMF-BRFL system was experimentally established using a 3 km PMF fiber. The laser can emit narrow-linewidth lasing light with a polarization state stably clamped onto one of the principal axes of the PMF, and the experimental results are consistent with the theoretical analysis. Furthermore, the effects of the pump power and cavity length on the working regions of the system and the characteristics of the lasing power, spectrum, and linewidth were investigated experimentally.

High-power laser facilities require a high-precision beam-target coupling, and one of its important error sources is the stability of the target system. There are inevitable internal vibration sources in a cryogenic target system that cause the vibration response of the slender cantilever structure, thus reducing the stability of the target system. The cryogenic target assembly is a slender cantilever-beam structure located at the head of the target system. Therefore, the stability of the suspension end of the target assembly significantly affects the accuracy of the beam-target coupling. There are two common methods of vibration control for cantilever beams, namely, active control and structural optimization. The active control method requires additional control circuits and driving mechanisms that can easily fail and have poor reliability in the low temperature and strong magnetic field environment of the vacuum target chamber. The latest Laser Megajoule (LMJ) device adopts the structural optimization method of reducing the length-to-diameter ratio of the target assembly. However, this method increases the mass of the cryogenic target head, and the stability optimization effect of the target system is not obvious. Compared to the less reliable active control method, the structural optimization method is worth further discussion. The structure specific-stiffness of the target assembly is improved, and the vibration response characteristics are optimized without changing the shape and mass of the original target cantilever beam. This method can effectively improve the stability of the target assembly and provide a reference for the design of target assemblies in future high-power laser facilities.

The National Ignition Facility (NIF) cryogenic target system is used as a reference. First, the vibration source is analyzed using relevant literature, and a 1∶1 target assembly model is built according to the data. The preliminary scheme design establishes a damping structure consisting of multiple sets of tensioning wires that contains two damping structure forms. In this scheme, a steel wire with an elastic damping property is attached to the middle of the cantilever beam of the target assembly, and the middle supporting point of the target assembly is added to change the vibration response characteristics of the target assembly. According to the design, the support for the spring damping Bernoulli-Euler equation of the cantilever beam is set up, the modal matrix is sorted based on the theory of mechanics of materials to establish formulas for calculating the stiffness of tensile steel wire vibration components, and the target assembly response function of the installation position of the vibration reduction component, diameter of the wire section, and three key research series damping size parameters are determined. The three theoretical parameters correspond to five parameters in an actual engineering structure: the number of components, installation position, diameter of the steel wire, value of the preload, and damping. ANSYS finite element analysis software simulates and optimizes these five structural parameters. The optimal value is used to design the vibration damping target assembly in detail, and a model of the assembly is constructed and assembled (Fig. 9). Two vibration sources can be accurately simulated: 1) an eccentric mass block and a direct current (DC) motor can simulate the vibration source of a refrigerator operation, and 2) a fixed track and heavy objects are used as the source of impact vibration when the insulation cover is opened. To characterize the optimization effect for the response amplitude and impact convergence time of the damping structure, the vibration damping target assembly with different parameters and design schemes is tested on a vibration test bench.

In the simulation, the total mass of the target assembly, natural frequency, and amplitude control effect are considered comprehensively (Fig. 6), and the installation position of the vibration damping structure is finally determined at the pressure plate. After the installation position is determined, the influence of the steel-wire diameter is simulated. We find that when the diameter is less than 0.8 mm, the vibration amplitude of the suspension end of the target assembly decreases uniformly as the diameter increases. When the diameter is greater than 0.8 mm, the vibration amplitude increases with the increase in diameter in an oscillatory manner (Fig. 7). This may be because when the diameter increases, the energy generated by the vibration of the vibration damping component also increases. When the critical point of 0.8 mm is reached, the vibration coupling effect between the vibration damping component and target assembly becomes evident, and the final amplitude increases in an oscillatory manner. The lowest point in the 0.8 mm diameter stationary zone, the point with a large amplitude in the 1.0 mm diameter oscillation zone, and the point with a small amplitude in the 1.2 mm diameter oscillation zone are selected for simulation. The theoretical optimal optimization rate of the damping target assembly is obtained as follows: the amplitude is 90% and the impact convergence time is 55% (Table 6). In the simulation experiment, the variation trend of the amplitude control effect of the two vibration damping structures agrees with the simulation results in Fig. 6, which proves that an oscillation zone exists and the minimum amplitude point of the oscillation zone can be realized by customizing the overall mass ratio of the steel wire. The optimal optimization rate of the damping target assembly measured in the simulation experiment is as follows: the amplitude optimization rate is 91.7% and the impact convergence time optimization rate is 77.1% (Table 8 and Fig. 13). Comparisons show that the integrated control effect of the designed series structure of the steel wire and damping material is superior (Tables 7 and 8).

In this study, a structural stability optimization design of the long cantilever structure of a cryogenic target assembly in the NIF is carried out, and a vibration-damping structure in the form of a vibration-damping component is designed according to the characteristics of lightweight and adjustable damping. Mathematical modeling shows that the control effect of the damping structure is mainly related to the installation position, diameter of the steel wire, and series damping. After the simulation experiment, the parameters are optimized, and the initial structural design is modified. In the simulation experiment, the experimental data on the vibration response characteristics of the target assembly are consistent with the simulation results. The results prove that the optimal control point of the damping structure in the oscillation region is achievable. The existing simulation experiments achieve the amplitude optimization rate of 91.7% and the impact convergence time optimization rate of 77.1%. Experiments show that the vibration-damping structure designed in this study has a good optimization effect on the vibration response amplitude and convergence time, and the wire and damping material series design scheme has a better comprehensive effect.

Continuous progress in laser processing technology and its growing industrial demand have resulted in short-wavelength blue lasers gradually becoming a research hotspot in the field of laser research. Blue semiconductor lasers have broad application prospects in precious-metal laser processing, laser-based cosmetic treatments, additive manufacturing, and other fields. Infrared lasers are usually used for metal processing in industry; however, owing to the high reflectivity of non-ferrous metals such as copper, gold, and aluminum in materials, the absorption effect of infrared wavelength lasers is low. In addition, conventional infrared lasers are bulky and complicated to operate and require high-power operation and complex cooling devices. The use of blue semiconductor lasers as a solution to process materials with high reflectivity and high thermal conductivity, such as pure copper, pure gold, and high-strength aluminum, has become a popular research topic in recent years. In addition, the spectral line width of a free-running blue light unit chip is usually 1 nm, which does not satisfy the requirements for spectral beam combination . Therefore, it is necessary to reduce the line width of blue light laser by technical means and simultaneously stabilize the output wavelength of the laser.

This paper proposes a blue laser with a narrow line width. First, we present the structural design of multiple single-tube blue semiconductor lasers. The design entails coupling multiple 447 nm blue light chips to form an optical fiber with core diameter of 105 μm and numerical aperture of 0.22 using spatial combination technology and the feasibility of this solution is verified by simulation using ZEMAX optical design software. Second, the laser line width is effectively narrowed using a reflective volume Bragg grating (RVBG). Because the output wavelength of each light-emitting unit of the free-running laser is different, the spectral line width of the output beam is increased. Therefore, the RVBG acts as an external cavity optical feedback element to enable the laser to output a single wavelength mode; in addition, the external cavity also serves to lock the wavelength. Finally, the narrow line width enables blue semiconductor lasers to deliver high-power performance, which can be detected from the optical path structure with the use of spectrometers. This lays the technical foundation for the practical realization of high-power blue lasers.

A photographic image of the output light source of the blue semiconductor laser is shown in Fig. 4. When the operating current is set to 3.0 A, the output spectrum of a single chip is stably locked at a wavelength centered at 444.07 nm after the light passes through the RVBG external cavity (Fig. 5). In terms of laser power, when the water-cooling temperature is 20 ℃, the threshold current of the free-running blue light chip is 0.6 A, and the six channels can output 1.26 W laser. After the addition of volume Bragg grating (VBG) external cavity feedback, the threshold is reduced to 0.5 A, and the six channels can output 1.38 W laser. Upon increasing the working current to 3.0 A, the output power is increased to 29.4 W after combining the laser beams. After RVBG external cavity feedback, the output power is 29.87 W, and the feedback efficiency reaches 101.6%. This is owing to the reduction in laser output threshold power after the addition of VBG external cavity feedback. In terms of spectral locking, multiple peak states exist in the spectrum before RVBG mode locking for a current of 3.0 A. After locking the RVBG mode, the mode-locking effect is clearly observed. The output is a single wavelength mode, the locked wavelength is 444.29 nm, and the spectral line width is narrowed to about 0.18 nm (Fig. 7). The module for narrow-line-width blue light coupling passes the power-current-voltage test. Under continuous conditions, the entire laser is adjusted within the driving current range of 0-3.0 A. When the operating current is increased to 3.0 A, the voltage is 25.1 V, and the output power of 26.32 W is obtained from the fiber with a core diameter of 105 μm and numerical aperture of 0.22. The electro-optical conversion efficiency is 34.95%, corresponding to a coupling efficiency of 88.1% (Fig. 8).

The RVBG is used as the feedback element to build a blue-light external-cavity semiconductor laser. Using spatial beam combination and fiber coupling technology, a laser output with a high power, narrow line width, and stable spectrum is obtained. The output power of 26.32 W is stable. The output wavelength is 444.29 nm, the spectral linewidth is narrowed to 0.18 nm, and the fiber coupling efficiency reaches 88.1%. Further experimental studies will be conducted to reduce coupling loss and improve spectral locking quality, and then combined with spectral beam combination technology, higher power blue semiconductor lasers will be obtained.

High-power narrow-linewidth fiber lasers are widely used in coherent synthesis and spectral synthesis; however, their power expansion is limited owing to stimulated Brillouin scattering. Common methods for inhibiting stimulated Brillouin scattering include the design and fabrication of stimulated Brillouin scattering suppression fibers and changing the temperature field and stress field distributions of the fibers. However, these methods entail complicated processing and can easily produce noise. In recent years, phase modulation of the light field has become the main method for suppressing stimulated Brillouin scattering. In the linewidth range of 50 GHz, single-stage phase modulation has a limited threshold boost for stimulated Brillouin scattering. In this study, we report a high-power narrow-linewidth fiber laser based on a cascaded pseudo-random binary sequence and sinusoidal phase modulation. The proposed method is expected to contribute to the power scaling amplification of narrow-linewidth fiber lasers in the 50-GHz linewidth range.

In this study, the appropriate pseudo-random binary sequence phase modulation parameters and low-pass-filter cutoff frequency are selected so that the unit spectral linewidth has the greatest suppression of stimulated Brillouin scattering. The effects of the modulation frequency and depth of sinusoidal phase modulation on the laser spectrum are studied. By changing the modulation frequency and depth, a spectral form with a fundamental frequency as high as the sideband of the ±1 level is obtained. After cascading the pseudo-random binary sequence and sinusoidal phase modulation, the spectrum shows a near-flat-top morphology, which exhibits good stimulated Brillouin scattering suppression. According to the theoretical research results, a high-power narrow-linewidth fiber laser based on cascaded phase modulation is constructed, and the output powers and stimulated Brillouin scattering thresholds are compared at root mean square (RMS) linewidths of 20 GHz and 46 GHz.

Based on previous research, when the pseudo-random binary sequence phase modulation depth is 0.55π and the ratio of filter cutoff frequency to modulation frequency is 0.53, the modulation spectrum exhibits a near-flat-top morphology (Fig. 2). According to theoretical research, the depth of sinusoidal modulation influences the number of spectral lines and the relative intensity of spectral lines, and the results are shown in Fig. 3. As the modulation depth increases, the number of spectral lines increases and the relative intensity of the spectral lines also changes. When the modulation frequency is reduced, the spectral line spacing and spectral linewidth are also significantly reduced. When the sinusoidal modulation frequency is 9.7 GHz and the modulation amplitude is 0.458π, the modulation spectrum has three single frequencies (Fig. 4). After cascading the pseudo-random binary sequence and sinusoidal phase modulation, the flatter spectral morphology than that from single-stage pseudo-random binary phase modulation is obtained, and the stimulated Brillouin scattering suppression is better. Based on the above research, a narrow-linewidth laser is built, and the output power and backward transmission power at RMS linewidths of 20 GHz and 46 GHz are monitored in the experiment (Fig. 7). When the RMS linewidth is 20 GHz and the output power is 2.2 kW, the backward power increases exponentially. When the RMS linewidth is 46 GHz,the output power is close to the SBS threshold. Because the stimulated Brillouin scattering phenomenon is not observed in the experiment, the power is increased to 4.93 kW, yielding a stimulated Brillouin scattering threshold enhancement factor of ~328, a system slope efficiency of 78%, and a beam quality factor (M²) below 1.2.

In this study, the physical mechanism by which the cascaded pseudo-random binary sequence and sinusoidal phase modulation is used to broaden the laser spectrum to suppress stimulated Brillouin scattering is investigated. The effects of pseudo-random binary sequence modulation frequency and mode length on the spectrum are theoretically studied. Under the optimal ratio of the filter cutoff frequency to the pseudo-random binary sequence modulation frequency, the unit linewidth well suppresses stimulated Brillouin scattering. Through theoretical simulations, the influence of the modulation depth and modulation frequency of the sinusoidal signal on the laser spectrum is obtained. Based on theoretical guidance, a narrow-linewidth single-fiber laser based on cascaded phase modulation is built for the experiment. The cascaded pseudo-random binary sequence and sinusoidal phase modulation is used to widen the seed source spectrum. Compared with the stimulated Brillouin scattering thresholds and output powers under different RMS linewidths, the output power finally reaches 4.93 kW after amplification by the four-stage optical fiber when the RMS linewidth of the seed source is 46 GHz. The system slope efficiency is 78% and the beam quality factor M² is below 1.2.

Solid-state lighting (SSL) has been widely used in indoor and outdoor lighting, including landscape lighting and automobile headlights. As mainstream SSL devices, phosphor-converted WLEDs (pc-WLEDs), exhibit many advantages such as low energy consumption and environmental friendliness. However, the applications of pc-WLEDs in high-power fields have been restricted because of their poor thermal stability, which originates from organic encapsulation. Glass ceramic phosphor (GC) is an important material required to resolve the problems resulting from high thermal stability by the means of remote encapsulation. Unfortunately, the chemical reaction between the red phosphors and glass powders during the co-sintering process significantly hinders the development of GCs. In this study, GCs are prepared by a screen-printing technique using a binder of borosilicate glass powder and a sapphire matrix with superior thermal conductivity. The phosphors used for the preparation of GCs are compounded with (Ca,Sr) AlSiN₃∶Eu²⁺ (CSASN) and Y₃Al₅O₁₂∶Ce³⁺ (YAG). The experimental processes of the GCs are simulated, and their optical properties are analyzed. Superior optical properties and color-tunable GCs are obtained. All results reveal that the as-prepared GCs are promising candidates for high-power white illumination.

GC preparation was divided into two parts. The first was the preparation of frit-seal glass, which was prepared by a rapid melt-quenching technique with a composition of SiO₂-B₂O₃-ZnO-Al₂O₃-CaO. One part of the glass melt was poured into a graphite mold and annealed to remove the thermal stress. After natural cooling, the sample was polished into a small cube for detection. The other part was poured directly into deionized water to quench the cullet, which was then dried and milled into glass powder. The second part was the preparation of the GC, using a screen-printing technique utilizing slurries, substrates, and screens. The slurries, consisting of CSASN, YAG, frit-seal glass powder, and an organic binder solution, were manually printed on the sapphire substrate using a screen. The printed samples with mass fractions of 17% YAG+(17%-x) CSASN+83% glass powders (where x=0%,1%,2%,3%, and 17%) were marked as Y-GC, Y-R1-GC, Y-R2-GC, Y-R3-GC, and R-GC series, respectively. After drying in an oven, the printed samples were sintered in a muffle furnace and allowed to cool naturally. Further, a series of GCs were prepared, and related characterization was performed.

The pattern of Y-GC corresponds well to that of YAG phosphor. Compared with the pattern of the CSASN phosphor, the intensities of some peaks of R-GC become higher or lower and some diffractions disappear, which originates from the preferential orientation when CSASN phosphors are prepared into R-GC (Fig. 2). The experimental GC processes are simulated using a DSC device. All GC samples show flat lines, similar to the results for glass powder (Fig. 3). As shown in the SEM images of the GCs (Fig. 4), the YAG and CSASN particles exhibit an acceptable homogeneous distribution in the glass, and the thickness of the samples is approximately 210 μm. The values of n_Y∶n_Al (atomic ratio) and n_Sr∶n_Al∶n_Si in the EDS spectra of the GCs correspond to those of the phosphors, and no additional elements appear in the glass regions. Meanwhile, distinct boundaries appear between the elements of CSASN and the glass powder in the EDS spectra. The above results, shown in Fig. 2 to Fig. 4, indicate that the two phosphors have no obvious chemical reactions with the sealing glass during the co-sintering process for GC preparations. With an increase in the red phosphor content, the PL spectra of the GCs exhibit a red shift (Fig. 5), and the intensities and quantum yields decrease gradually (Fig. 6). The results are ascribed to the changes in the microstructure of CSASN during GC preparation. The thermal stability of the GCs decreases gradually, but the rate of decline is small. The PL intensity at 150 ℃ is maintained at 89% of the initial value (Fig. 7), which is attributed to the high thermal conductivity of the frit-seal glass and sapphire substrate. With the increase in red phosphor content, the GC samples show a color change from yellow to red and a color-tunability from cool white to warm white encapsulated with a blue chip (Fig. 8). GCs exhibit a decrease in luminous efficiency (LE), a decline in correlated color temperature (CCT), and an increase in the color rendering index (CRI). The results are due to the small luminous flux of red light and the QY decrease in the GC. The LE gradually decreases as the operation current increases from 20 to 150 mA, which is ascribed to the phenomenon of "efficiency droop" (Table 1).

GCs containing YAG and CSASN phosphors are prepared using a screen-printing technique on a sapphire substrate. The results indicate that the two phosphors have no obvious chemical reactions with the sealing glass during the co-sintering process for the GC preparation. With increasing red phosphor content, the PL spectra exhibit a red shift. Meanwhile, the intensities and quantum yields gradually decrease. The GCs show superior thermal stability, and the PL intensity at 150 ℃ is maintained at 89% of the initial value. The GC samples exhibit color tunability from cool white to warm white with increasing red phosphor content. The optimal photoelectronic parameters of the GCs are a luminous efficiency of 147.70 lm/W, color temperature of 4915 K, and color rendering index of 73.3. These results indicate that the as-prepared GCs have a wide range of applications in the field of high-power white illumination.

Heterodyne laser interferometers have been widely utilized in the field of precision measurement owing to their wide measurement range, high measurement accuracy, and robust measurement ability. With the development of high-precision science and technology, higher requirements have been proposed for the measurement accuracy of heterodyne laser interferometers, which is significantly affected by the phase measurement accuracy of the interference signal. In multichannel signal acquisition and processing, the crosstalk error between sampled signals is a familiar error source. In this study, the problem of signal crosstalk in interference signal processing is investigated, and a signal crosstalk error model is established. Accordingly, a pre-compensation method based on spectrum analysis is proposed to eliminate signal crosstalk errors and improve the phase measurement accuracy of the interference signal.

The signal crosstalk error was systematically examined through theoretical derivation, simulation analysis, and experimental verification to solve the problem of crosstalk between two sampled signals in heterodyne interferometry. First, a mathematical model of the signal crosstalk error was deduced, and a pre-compensation method based on spectrum analysis was proposed. Then, the influence of the crosstalk coefficient, signal amplitude ratio, and crosstalk signal phase offset on the signal crosstalk error was analyzed via simulation, and the crosstalk compensation method was verified. After the initial verification of the error model and compensation method through simulation, the signal processing algorithm was implemented based on the Red Pitaya FPGA board, and further verification was performed through a phase measurement experiment. The experimental results show that the actual measurement error is consistent with the theoretical calculation error and that the crosstalk compensation method can effectively eliminate the signal crosstalk error. Finally, a heterodyne interferometric measurement system was built to verify that the proposed signal processing system can meet the measurement requirements of practical applications.

This study deduces a mathematical model of the signal crosstalk error and proposes a pre-compensation method based on spectrum analysis. Then, the analysis and verification are conducted through simulations and experiments. According to the derived mathematical model of the signal crosstalk error, the crosstalk coefficient, signal amplitude ratio, and crosstalk signal phase offset affect the magnitude of the crosstalk error. The influence of these three factors on the signal crosstalk error is analyzed via simulation. The crosstalk coefficient and signal amplitude ratio significantly impact the size of the signal crosstalk error (Figs. 3 and 4), and the phase offset of the crosstalk signal affects the size of the crosstalk signal error and the location of the extreme value distribution simultaneously (Fig. 5). After the simulation analysis, the signal processing algorithm is implemented based on the Red Pitaya board, and a phase measurement experiment is performed (Fig. 8). When the amplitude ratios of the two input signals are 1, 2, and 3, the actual measurement and theoretical calculation errors are consistent (Fig. 13). After compensating for the signal crosstalk error, the maximum measurement error drops from 0.34° to 0.01° (Fig. 14). A phase measurement experiment verifies the correctness and effectiveness of the signal crosstalk error model and compensation method. Finally, a heterodyne interferometric measurement system is built to test the performance of the algorithm (Fig. 16). In the range of 250 μm, the measurement error is less than 5 nm (Fig. 17), indicating that the signal processing algorithm can meet the needs of actual measurements.

Phase measurement accuracy is essential for accurate heterodyne laser interferometer measurements, and the signal crosstalk error is a common source in multichannel signal sampling processing. This study deduces the signal crosstalk error model and proposes a pre-compensation method based on spectrum analysis to solve the problem of crosstalk between the reference signal and measurement signal in heterodyne interferometry. The influence of the crosstalk coefficient, signal amplitude ratio, and crosstalk signal phase offset on the signal crosstalk error is analyzed via simulation. When both crosstalk coefficients are 0.01, the signal amplitude ratio is 10, and the phase offset of the crosstalk signal is 0, the maximum crosstalk error of the signal can reach 5.78°, which needs to be effectively compensated. In the phase measurement experiment, when the amplitude ratios of the two signals are 1, 2, and 3, the actual measurement error is the same as the theoretical calculation error, proving that the signal crosstalk error model is correct. After the signal crosstalk error compensation, the measurement error drops from the maximum of 0.34° to 0.01°, proving that the crosstalk error compensation method is effective. In summary, this study systematically analyzes the signal crosstalk error. The proposed compensation method can effectively eliminate the signal crosstalk error and improve the phase-measurement accuracy of the heterodyne interference signal.

Currently, high-reflectivity optical elements are widely used in optical fields, such as ring laser gyroscopes, inertial confinement fusion systems, and gravitational wave detection. Their reflectivity values have a decisive influence on the performance improvement of these systems; therefore, accurate measurement of their reflectivity values is necessary to optimize the system performance. At present, there are many measurement methods, among which spectrophotometry and cavity ring-down (CRD) technology are the most widely used. However, these two methods have certain shortcomings. Spectrophotometry is associated with low accuracy, and its measurement limit is usually lower than 99.9%; especially when measuring high reflectivity (>99%), its accuracy is reduced. Devices using CRD technology for high reflectivity measurement are relatively complex and expensive, and the measurement conditions, such as adjusting the incident angle and polarization state, are inflexible. Moreover, according to the measurement principle, the lower the reflectivity of the sample to be measured, the lower is the measurement accuracy. Therefore, for samples with the reflectivity of 99.9%-99.99%, existing testing methods have some limitations. Based on the described situation, it is necessary to study a high-precision reflectivity measuring device that can measure the reflectivity of 99.9%-99.99% accurately. It is necessary to develop a reflectivity measuring device which has the advantages of simple device, convenient adjustment and good measurement stability.

Based on traditional spectrophotometry, we study a more accurate reflectivity measurement method. The main improvement in this method is measuring the difference signal instead of the reflected light and incident light signal. In this study, a reflectance measuring device based on spectrophotometry is developed, which adopts a double-path measurement. First, we measure the signal difference between the reference optical path and the initial optical path without samples, and the reference optical path signal (Fig. 2). We then measure the signal difference between the reference optical path and the test optical path after placing the sample (Fig. 3). The reflectivity of the sample is calculated according to the measured reference and differential signals. The reflectivity is calculated by measuring the difference between the reference signal and the initial signal and the difference between the reference signal and the test signal. Compared with the reference signal, initial signal, and test signal with larger absolute values, the signal difference itself is relatively small; therefore, the sensitivity of the lock-in amplifier can be fully utilized to improve the reflectivity measurement accuracy. In addition, we use a quick fixing mechanism to shorten the measurement time and reduce the influence of unstable factors such as the light source and environment on the experimental results. In the experiment, there are some differences in the responses of the detectors' photosensitive devices at different positions; therefore, mobile platforms are installed at the three positions where the two detectors are placed. Before the formal measurement, the position of the detector is adjusted to ensure that the beam irradiates the position where the conversion efficiency is the highest on the receiving surface of the detector .

We use this device and the CRD method to test the highly reflective mirror samples. The wavelength in the two methods is 1064 nm, and the incident angle is 45°. The high-reflection wavelength of the two high-reflection mirror samples is 1064 nm, the use angle is 45°, the lens diameter is 50 mm, and the thickness is 5 mm. The two samples are measured using S-polarized light and P-polarized light, respectively, and the highest reflectivity is 99.986% (Table 1). In this study, the error is calculated using an uncertainty transfer formula. When only one significant digit is retained, the error is of the order of 10^-5,so the reflectivity calculation result is also of the same order, which meets the reflectivity measurement requirements of a high reflector. The accuracy of the measurement method introduced in this study reaches 0.01%. Compared with the CRD method, the measurement error is less than 0.009%, which shows that this method can achieve a higher measurement accuracy with a simpler device. The advantages of this device are as follows: 1) By using a lock-in amplifier to measure the signal with high precision and a small range and shortening the measurement time, higher accuracy can be achieved; 2) The sample fixture is installed on the rotating platform, and the adjustable position range of the detector of the test optical path is large, so the adjustable incident angle range of the device is also large; 3) The device is simple and easy to operate.

In this study, traditional spectrophotometry is improved, the measurement accuracy is increased by measuring differential signals, and the high-precision range of the lock-in amplifier is fully utilized. In addition, in the calibration of the device, the measurement time is shortened by means of a quick fixing mechanism, and the influence of light source fluctuation is reduced, which ensures the accuracy of the measurement results. The measurement accuracy can reach 0.01% stably, which meets the measurement requirements of the existing high-reflector elements, compensates for the shortcomings of the traditional measuring methods, greatly increases the application range of spectrophotometry, and fills the measuring gap between commercial spectrophotometry and CRD technology in the reflectivity range of 99.9%-99.99%.

As an indispensable visibility ground-based observation instrument in atmospheric environment monitoring, the forward scatterometer currently adopts the standard scatterer external field calibration method, that is, multiple standard scatterers with different scattering characteristics are placed in the working light path in turn, so as to transfer the measured values of the forward scatterometer to the scattering characteristics calibration results of the standard scatterer. For the calibration of standard scatterers, the calibration method based on the field comparison test between the forward scatterometer and the atmospheric transmission meter is widely used internationally at present, which can trace the calibration results of standard scatterers to the calibration accuracy of the neutral density attenuator used to calibrate the atmospheric transmission meter. However, this method has complex process, many error transmission links, serious influence from weather conditions, long calibration time, and low efficiency. Moreover, the calibrated standard scatterer can only be used for the field calibration of a specific type of forward scatterometer, which has poor universality. At the same time, the field calibration results of the forward scatterometer are subject to the traceability chain of the measured values of the atmospheric transmittance, and because the calibration light sources for calibrating the neutral density attenuator of the atmospheric transmittance are mostly tungsten halogen lamps, xenon lamps and monochrome or white LEDs, the spectral distribution is different from that of the 2700 K color temperature incandescent lamps in the definition of meteorological optical range (MOR), the field calibration results of the forward scatterometer cannot be traced to the definition of MOR. In this paper, a calibration method of standard scatterers traceable to the definition of MOR is proposed, and a calibration optical system of standard scatterers used to calibrate the forward scattering visibility meter is designed. The synchronous calibration of multiple scattering angles is realized, the calibration time is greatly shortened, and the universality is strong. At the same time, the problem that the calibration of standard scatterers cannot be traceable to the definition of MOR is solved, providing a technical basis for improving the transmission and traceability system of visibility values.

According to the standard scatterer calibration method, a standard scatterer multi-angle synchronous calibration optical system architecture is determined, and a standard scatterer calibration optical system is designed for calibrating the forward scattering visibility meter. The standard scatterer calibration system is divided into an illumination optical system and a light field measurement optical system. The illumination optical system includes a light source, a shaping lens, a double-row fly-eye lens and a collimating lens group. The optical system for light field measurement is mainly composed of standard scatterer, hemispherical dome, aspheric reflector, panoramic imaging system and charge-coupled device (CCD) cameras. The illumination optical system and light field measurement optical system are optimized, and a calibration system energy calibration method is studied. The energy calibration error of the calibration system is simulated using standard Lambert scatterers.

The spectral distribution simulation based on the ideal 2700 K color temperature incandescent lamp verifies the calibration accuracy of the standard scatterer calibration method traceable to the definition of MOR. The designed standard scatterer calibration optical system realizes multi-angle synchronous calibration with pitch angle of 20°-50° and azimuth angle of 0°-360° (Fig. 4). The study on the energy calibration method of the calibration system shows that when the standard scatterer calibration system conducts energy calibration within the range of 20°-50° of the scattering angle, the maximum circumferential average relative error is 2.28% when the scattering angle is 31° (Fig. 15), which proves the improvement in calibration accuracy of the standard scatterer. The simulation verification shows that the magnitude range of MOR represented by the standard Lambert scattering angle of 20°-50° is 6.54-45.74 m, which conforms to the definition of MOR by the World Meteorological Organization, and the maximum absolute error is -2.41 m (Fig. 16).

Using the standard scatterer calibration method that can be traced back to the definition of MOR, a standard scatterer calibration optical system used to calibrate the forward scattering visibility meter is successfully designed, and the multi-angle synchronous calibration with pitch angle of 20°-50° and azimuth angle of 0°-360° is realized. The maximum absolute error of the standard scatterer calibrated by the calibration method is better than 1/10 of the measurement accuracy requirement of the forward scatterometer specified by International Civil Aviation Organization (ICAO), providing theoretical basis and technical support for the calibration of the standard scatterer traceable to the definition of MOR.

Subwavelength optical field confinement and low-loss propagation are important for compact photonic integration. However, the field confinement capability of noble metal-based plasmonic devices is always accompanied by the inherent ohmic loss. These structures perform well at the near-infrared and visible frequencies. However, in the mid- and far-infrared regions, the lack of tunability in their electromagnetic response and poor modal field confinement ability hinder their applications at the nanoscale. Recently, both theoretical and experimental reports have demonstrated that graphene can support surface plasmons in the infrared range with unique tunability, extremely strong modal field confinement, and a large field enhancement. Owing to the tradeoff between modal loss and confinement in plasmonic structures, it is difficult to obtain a deep-subwavelength modal field and long-range propagation simultaneously. Although a graphene-based hybrid waveguide integrated with a triangle wedge substrate has been proposed to achieve an ultra-small normalized mode size, the modal loss remains relatively large (with a propagation length less than 10 μm). Thus, obtaining a better balance between the modal loss and field confinement remains a significant challenge. In this study, a hybrid plasmon waveguide consisting of a cylindrical silicon nanowire and graphene-coated triangular nanowire is designed. The designed waveguide exhibits strong optical field confinement capability, low loss propagation, and a high figure of merit; thus, it is suitable as a building block for subwavelength photonic devices, such as modulators and nanolasers.

The proposed plasmon waveguide consists of a graphene-coated triangular nanowire separated from a cylindrical silicon nanowire by a nanoscale dielectric gap with width (h_gap)(Fig. 1). For the modeling and simulation, the wave optics module of COMSOL software is employed. The eigenvalue solver is used to obtain the complex effective mode index (N_eff) and effective mode area (A_eff). In the simulation, the graphene layer is modeled as an electric field-induced surface current J=σ_gE without thickness, where J is the surface current and E is the electric field. The calculation domain is 2λ₀×2λ₀, and a perfectly matched layer (PML) is applied to the surroundings of the geometry to avoid the influence of reflection. A convergence analysis is performed to ensure that the numerical boundaries and meshing do not interfere with the solutions.

The proposed waveguide exhibits well-confined modal fields with the focal spot area of 13.5×10 nm² (even less), corresponding to size of λ02/(7.4×10⁵). The modal transmission properties are highly dependent on angle θ. When θ decreases, both the propagation loss and modal area decrease, indicating that the tradeoff between modal loss and confinement is broken to some extent. Additionally, a smaller gap distance results in a larger figure of merit (Fig. 3). Interestingly, when the radius of the Si nanowire increases, the propagation length decreases slightly at first and then appears to be invariable, while the modal area decreases monotonically. This is because the loss is mainly from the graphene layer, and when R increases, the graphene-light interaction region is nearly invariable. When R ranges from 40 nm to 100 nm, the loss is invariable, while the modal area reduces (Fig. 4). When the triangular nanowire permittivity ranges from 2 to 12, the figure of merit decreases from 1190 to 245, implying that a smaller permittivity can improve waveguide performance (Fig. 5). The tunability of the Fermi energy of graphene allows the active tuning of the modal properties. When the Fermi energy ranges from 0.4 eV to 1.4 eV, the normalized modal areas of the fundamental modes are on the order of ~10^-6. In particular, when the Fermi energy is above 1 eV, a propagation length of several tens of micrometers and a figure of merit of over 2500 can be achieved (Fig. 6). Finally, the proposed waveguide is demonstrated to have better comprehensive performance than a similar waveguide structure (Fig. 8) and a two-wire system (Fig. 9).

In this work, the subwavelength transmission characteristics of a graphene-dielectric nanowire hybrid waveguide are proposed and examined. The proposed waveguide exhibits extremely strong optical field confinement capability. In particular, when θ decreases, both the modal loss and modal area decrease. This is due to the tip focusing effect of the triangular-shaped GNW, and the decrease in θ results in a reduction in the graphene-light interaction region, which in turn reduces the loss. Thus, the tradeoff between modal loss and confinement is broken to some extent. When the frequency ranges from 20 THz to 40 THz, the normalized modal areas of the fundamental modes in the designed waveguide are on the order of ~10^-6 with a propagation length of approximately several tens of micrometers, as well as a high figure of merit of over 2500. Compared with a similar waveguide structure, the mode field area is reduced by one order of magnitude while maintaining a comparable propagation distance. These findings are expected to have potential applications in waveguide-integrated plasmonic devices and greatly reduce the device size, such as modulators and nano-lasers.

Nanostructures based on metallic materials can modulate the amplitude, phase, and polarization of electromagnetic waves owing to their surface plasmon resonance (SPR) properties. The interference between bright and dark modes forms Fano resonances in metamaterials. Excitation of the dark mode can effectively suppress far-field radiation and enhance near-field radiation. However, the significant heat loss of metallic materials limits their application in optics; therefore, only a few superconfigurable materials based on surface plasma excitations can be used in practical applications. Recent studies have shown that highly refractive index all-dielectric nanostructures with low absorption properties do not undergo heat loss, thus facilitating the realization of high-performance compact devices. In this study, we designed a fully dielectric nanopillar supersurface with a high Fano resonance quality factor, Q, and modulation depth. We hope our design can provide innovative ideas for asymmetric transmission, polarization angle detection, and super-surface multifunctional multiplexing.

In this study, the Fano resonance theory was simulated around a fully dielectric supersurface material. Maxwell' s equations describe the electromagnetic-wave propagation law in space, and the equations can be solved to determine the response of the supersurface to the incident light. However, the analytical solution of Maxwell' s equations cannot be obtained in general; therefore, the simulation results are typically obtained by solving a system of equations using numerical methods. The two widely used solution methods are the finite element method (FEM) and the finite difference in the time-domain method (FDTD). We used the FDTD Solutions software to simulate the supersurface and perform high-precision simulations to replace the more expensive prototype experiments. The periodic boundary conditions were set in the x- and y-directions owing to the periodicity of the superlattice structure, and a perfect matching layer (PML) was set in the z-direction. In addition, the polarization plane wave was vertically incident in the negative direction of the axis. Simulations were performed sequentially by changing the nanopillar structure to analyze the Fano resonance generation mechanism.

The designed full-dielectric supersurface has a high-quality factor, Q, and modulation depth. Flexible modulation from single-Fano resonance to double-Fano resonance can be achieved by increasing the number of nanocolumn rows. The transmission spectrum of the first simulated single-row nanocolumn and the electromagnetic field distribution show that the Fano resonance (Fig. 3) was generated by a toroidal dipole but with a decreased quality factor. The coupling between the nanocolumns can be modulated by increasing the number of nanorows such that the toroidal dipole (TD) and magnetic dipole (MD) jointly dominate the dark mode, thus increasing the quality factor and enhancing the near-field coupling (Fig. 6). The final increase to the three rows of nanopillars achieves a double-Fano resonance. The first Fano resonance peak is formed by the TD and electric dipole (ED) resonance when the scattering power values are equal, and both interfere to cancel out each other to produce a radiation-free anapole mode. The second Fano resonance peak is formed by the resonant interference of the TD and MD to form the dark mode, whereas the remaining resonant modes act in the dark mode. The interference between the two modes forms the Fano resonance peak (Fig. 9). The sensitivity of methane volume fraction and the background refractive index can be measured simultaneously, and the simulation calculations show that the sensor has a high sensitivity (Fig. 12).

Based on the high Fano resonance quality factor, Q,of the Fano resonance metasurface and the local characteristics of the electric field, this study designed a fully dielectric structure and investigated the relationship between the hypersurface structure parameters and the quality factor Q and modulation depth. Effective control of the single-Fano resonance to double-Fano resonance is achieved via essential parameter optimization. By comparing the sensitivities of the rectangular column and elliptical cylindrical metasurface structures with the same parameters, the rectangular column structure is selected, and its optimal parameters are determined. The quality factor of the rectangular column metasurface structure reaches 3408, and the modulation depth of the double resonance peak is close to 99%. Through simulations, CH₄ volume fraction sensitivity values can reach 1.57 nm/% (for dip1) and 1.66 nm/% (for dip2). The background refractive index sensitivities are 419.45 and 395.7 nm/RIU, and thefigure of merit (FOM) values are 524.3 and 542.8 RIU^-1, respectively. Knowledge of linear algebra proves that the sensitivity error of the sensor is slight. In addition, the sensor can measure the deflection angle according to the magnitude of the resonance peak, and the manufacturing error tolerance of the sensor did not exceed 1.3 nm.

Photon counting LiDAR is widely used in target ranging, three-dimensional imaging, and other fields, owing to the advantage of high sensitivity. The return echo data are obtained in the photon counting LiDAR by recording the presence or absence of photon events at the corresponding time, resulting in the inability to acquire the target echo waveform in one detection. The cumulative histogram of the photon count is obtained by the accumulation of multiple detections. The probability histogram of the photon count is regarded as the photon return detection probability waveform, which is closely related to the true return waveform of the target. In traditional LiDAR, the distance of a target can be determined by calculating the centroid of the return signal. However, in photon counting radar, the detection probability waveform of the photon echo is significantly distorted relative to the target waveform owing to the long response dead time of the detector, which significantly affects the accuracy of the photon ranging and the effective acquisition of the target information. Most researchers recover photon echo information based on the detection probability function with a large data error under low signal-to-noise ratio (SNR), making it difficult to obtain the target echo waveform information. Therefore, we discuss the photon echo correction method for the photon counting signal with a low SNR in this paper.

A photon detection echo model is discussed based on the LiDAR detection equation and probability response of photon detection. Combined with the simulated annealing algorithm, the particle swarm optimization algorithm is modified to estimate the photon echo parameters, including the echo signal strength, signal pulse width, peak position of the signal, and average photon noise intensity. The simulated annealing algorithm makes the swarm particles jump out of the local optimal position and effectively improves the global solution search ability. However, to avoid losing the possible dominant particle population, only a few particles are randomly selected for simulated annealing. The consistency between the recovery signal and target true return signal is evaluated by defining the evaluation function. The algorithm’s accuracy is evaluated by calculating the difference between the real target location and location information determined by peak method for the recovered target echo signal.

The algorithm proposed in this study can achieve fine signal recovery results at a low SNR, whereas the iterative solution based on the photon detection probability has a severe signal recovery distortion (Fig. 2). When the total number of noise photons increases from 0.5 to 5.0, the difference between the recovered signal recovered by the iterative method and true signal increases from 0.007 to 0.061, and the ranging error oscillates from 5.5 cm to 13.7 cm. The difference between the recovered signal acquired by our algorithm and the real signal is always below 0.005, and the ranging error is below 3.4 cm (Fig. 3). The signal recovered by our algorithm, which still maintains good performance in the case of high noise, is closer to the real target echo signal. The photon detection experiments are conducted on a deep plane target with a distance of 120 cm from the front to the back. Using our method, the recovered distance of the two target signals is 123.45 cm and the error is 3.45 cm. The corresponding distance of the two target signal peaks obtained by the iterative recovery algorithm is 130.32 cm and the error is 10.32 cm (Fig. 6). The method proposed in this study can better extract the target information with the depth structure.

The simulation and experimental results show that the target signal recovery algorithm based on partial annealing particle swarm optimization can obtain stable target echo signal recovery results under the condition of a low SNR. Compared with the existing iterative method, it improves the effectiveness of signal recovery under the condition of a low SNR and avoids the increase of the error caused by iterative accumulation. Furthermore, the algorithm proposed in this paper has better performance in recovering the depth information of the target.

Precise perception of the surrounding environment is the basis for realizing various functions in autonomous driving. The accurate identification of the location of 3D targets in real scenes is key to improving the overall performance of autonomous driving. Lidar has become pivotal in this field because of its superiority in sensing richer 3D spatial information while being less affected by weather and other environmental factors. Current 3D target detection methods are mainly based on deep learning, which can achieve a higher detection accuracy than traditional clustering and segmentation algorithms. The key to target detection based on deep learning is the in-depth extraction and utilization of point-cloud feature information. If feature information cannot be fully utilized, the target is misdetected or missed (Fig. 1), which has a significant impact on the safety of the automatic driving function. Therefore, deep extraction and utilization of point cloud information are key to improving the accuracy of 3D target detection.

This study proposes a two-stage 3D target detection network (DSPF-RCNN, Fig. 1). In the first stage, the unordered original point cloud is divided into the regular voxel space, and the point-wise feature is converted into voxel-wise feature by using convolution neural network. The down-sampling output of the last layer is transformed into a 2D bird's eye view (BEV), whereby the BEV is input into the deep feature extraction-region proposal network (DFE-RPN, Fig. 2) for depth extraction of 2D features. Through the fusion of deep and shallow texture features with deep semantic features, the ability of the network to capture 2D image features is enhanced. In the second stage, some point clouds are selected as center points in the latter two 3D down-sampling voxel spaces through the farthest point sampling, and the center points are input into the aware-point semantics and position feature fusion (ASPF) module (Fig. 3), allowing the integration of the 3D semantic features and location information of the surrounding point clouds. In this manner, the network can adaptively extract more diverse features of the target because these center points have a stronger feature aggregation ability when aggregating neighboring point clouds, which improves the network's ability to aggregate different feature information of the target. These center points are then used to aggregate the features of the surrounding point clouds in the 3D voxel space (Fig. 4). Subsequently, the region-of-interest pooling is conducted for the aggregated features and target candidate boxes generated in the first stage. Finally, the more refined classification and boundary box regression are conducted for the target through the fully connected layer.

The DSPF-RCNN is tested and evaluated using the official KITTI test and validation sets. The detection results for Car are better than those of the existing mainstream algorithms in the test set (Table 1), and the detection accuracies at the three difficulty levels are 89.90%, 81.04%, and 76.45%. In the KITTI validation set (Table 2), at the 11 recall positions, the detection accuracy is improved by 4% compared with those of the SVGA-Net and Part-A2 networks at moderate levels for Car and Cyclist. The DSPF-RCNN can accurately detect the three types of targets (Fig. 5). The effectiveness of the proposed innovation module is further compared and analyzed (Table 5). The results show that, after integrating the 3D semantic features and position features of the surrounding point cloud, the central point can better aggregate the feature information of the surrounding point cloud in the feature aggregation stage. However, when the DFE-RPN module is added, the network's ability to capture features increase further, and the ability to extract small-target feature information, such as cyclists and pedestrians, is significantly improved. Finally, a comparative analysis is performed on the network time utilization, including the time consumed by each module in reasoning through a frame of point cloud data (Table 6). The comparison between DSPF-RCNN and the other two-stage algorithms (Table 7) shows that the total inference time of DSPF-RCNN is 64 ms, which is more advantageous in terms of the inference speed of the two-stage algorithm. Finally, the algorithm is deployed on a real vehicle platform to realize online detection (Fig. 7).

In this study, a two-stage target detection algorithm, the DSPF-RCNN, based on a laser point cloud is proposed. First, the proposed DFE-RPN module extracts abundant target feature information from 2D images. In the second stage, the proposed ASPF module allows the central points to aggregate the salient features of different targets. Through testing on the KITTI test set and validation set, and comparison with mainstream methods, it is concluded that DSPF-RCNN performance is more advantageous in accurately detecting targets with different sizes, including small targets. At moderate levels in the KITTI validation set, the detection accuracies for Car and Cyclist are improved by approximately 4%, and the total network inference time is 64 ms. Finally, the DSPF-RCNN is applied to a local dataset to verify its engineering value.

Recent advancements in laser scanners and photogrammetry technology have significantly reduced the cost of acquiring 3D point clouds. Consequently, various types of point clouds have gradually become popular data sources for urban applications. The accurate registration of cross-source and multi-temporal point clouds must be ensured before developing applications based on 3D point clouds. However, this is a challenging task owing to (1) the large amount of data to be considered, (2) the wide discrepancy in characteristics between cross-source point clouds, and (3) the significant changes in a scene represented by multi-temporal point clouds. These data characteristics can harm the extraction and matching of registration primitives, resulting in the poor performance of marker-free registration techniques. In this paper, we propose an automated, efficient, and marker-free method for registering cross-source and multi-temporal point clouds in urban areas.

The proposed registration method comprises three stages keypoint generation, correspondence matching, and transformation estimation. (1) Keypoint generation. We generate object-level virtual keypoints as registration primitives rather than directly extracting local features from point clouds, which are redundant and sensitive to outliers and missing data. Specifically, the ground points are first filtered out via the cloth simulation filtering algorithm. The remaining points are decomposed into planar segments by fitting planes in a region-growing manner. Finally, virtual keypoints are determined as the endpoints of intersecting line segments of two adjacent planes. (2) Correspondence matching. First, local triangles are constructed using the generated virtual keypoints as vertices to encode the relative spatial relationships among keypoints within a point cloud. Second, the triangle sets of both point clouds are mapped to a feature space where the triangles become 3D feature points. For each feature point in the source point cloud, we determine its closest point in the target point cloud, forming triangle pairs between the two point clouds. Finally, we propose an improved global matching approach with linear time complexity to extract correspondences encoded in the triangle pairs. (3) Transformation estimation. As cross-source and multi-temporal point clouds are typically well-leveled, registration can be achieved by aligning the two point clouds horizontally and translating them vertically. We use the horizontal coordinates of the correspondences to estimate the 2D horizontal transformation and their vertical coordinates to calculate the vertical translation.

We evaluated the effectiveness of the proposed method using large-scale real-world urban point clouds. The experimental data consist of six cross-source and multi-temporal point clouds, including three airborne light detection and ranging (LiDAR) point clouds and three photogrammetric point clouds, which cover an urban area of 1.8 km² in Rotterdam, the Netherlands. Each point cloud comprises a large number of points (approximately 20-60 million points per point cloud; refer to Table 1 for details). Additionally, as the point clouds were collected over a long period of time, many of the objects in the scene have changed considerably. These two characteristics make them suitable for performing comprehensive evaluations of automatic marker-free registration methods. To evaluate the registration results qualitatively, we visualized a randomly selected region (Fig. 7) and three manually selected buildings with varying architectural styles (Fig. 8). Despite the different characteristics of cross-source point clouds and the significant changes in scenes, the proposed method could accurately align all five registration pairs formed by the six experimental point clouds. To evaluate the registration results quantitatively, we calculated both matrix-based errors (i.e., rotation and translation errors) as well as pointwise errors. The evaluation is summarized in Table 4. Our automatic registration results have an average pointwise error of 6.4 cm, whereas the average matrix-based errors are 0.2′ for rotation and 7.4 cm for translation. Furthermore, despite the massive size of the experimental point clouds, the proposed approach required only 105.7 s to achieve pairwise registration on average. Both qualitative and quantitative results demonstrate the effectiveness of the proposed method for registering cross-source and multi-temporal urban point clouds.

A fully automated marker-free registration approach is presented for cross-source and multi-temporal point clouds in urban environments. Object-level virtual keypoints are generated from urban point clouds as registration primitives, thereby overcoming the challenge of identifying valid corresponding features. By encoding rigid body spatial relations among the generated virtual keypoints, we establish correspondences between the source and target point clouds, resulting in efficient matching for large-scale urban scenes. Experiments on real-world data demonstrate that the proposed method can automatically, accurately, and efficiently register cross-source and multi-temporal point clouds in urban areas, indicating its practical utility. In the future, we would like to collect more data to test the robustness of the proposed method. Moreover, we intend to study the potential of the proposed matching algorithm in the fusion of general multi-source data, e.g., aligning 3D building point clouds with 2D building footprints.

Fiber Bragg grating (FBG) is advantageous owing to its compact size, light weight, anti-electromagnetic interference, high-temperature resistance, and series multiplexing, and it is increasingly adopted in several fields such as aerospace, petrochemical industry, national defense, and military. The measurement and analysis of FBG reflection or transmission spectra yield the magnitude of the physical parameter to be measured (Fig.1). Extracting characteristic information from FBG spectral signals, i.e. FBG signal demodulation, is the basis for FBG sensing measurement applications. Most of the current FBG sensing-based systems measure at frequency below 1 kHz and are mainly used in the measurement of slowly changing physical quantities such as temperature and strain. For scenarios requiring high-speed dynamic measurements, such as high-speed vibration and explosive shock, the signal demodulation scheme based on FBG sensing must satisfy the measurement speed requirements for effective application. Depending on the measurement parameters, the FBG signal high-speed demodulation methods can be divided into four categories: spectral, optical-intensity, phase, and microwave spectrum analyses (Fig.2).

The spectral analysis method (Section 3) relies on the full FBG spectrum for measurement analysis, which can be achieved via dispersive and scanning spectroscopies. Spectral measurement based on the principle of dispersion (Section 3.1) can be divided into spatial- and time-dispersion methods. Spectral measurement using the principle of spatial dispersion is the main method for FBG demodulation. The test speed depends on the scanning speed of the spectral acquisition device such as the charged-couple device (CCD). Spectral measurement systems using the principle of time dispersion require ultra-high-speed acquisition equipment. To obtain sufficient time delay, the length of the dispersion element is quite long. Hence, reducing the size of the dispersion element and guaranteeing a sufficient amount of dispersion is the key to its application. Using tunable light sources such as distributed Bragg reflector (DBR), distributed-feedback laser (DFB), Fourier domain mode-locked laser (FDML), and high-speed detectors can also obtain FBG characteristic spectrum, i.e., scanning spectroscopy (Section 3.2). The scanning speed of the tunable light source is the main factor that determines the testing speed of the system. Light intensity analysis (Section 4) can be achieved by single- (Section 4.1) or double-edge filtering (Section 4.2). The corresponding measurement system has no mechanical structure, and the measurement speed can easily reach the megahertz level. Its measurement range, sensitivity, and linearity are determined by the performance of the filter device. Presently, the optical filter components available are long-period fiber grating (LPFG), array waveguide grating (AWG), Fabry-Perot (FP) cavity, etc. The phase analysis method (Section 5) demodulates the interference signal by measuring its phase. Common systems include the Mach-Zehnder (Section 5.1), Michelson (Section 5.2), and Sagnac (Section 5.3) interference structures. Although precision is high, the demodulation range is small, and demodulation speed of several hundred kilohertz can be achieved. Microwave spectrum analysis (Section 6) converts optical domain signals into microwave frequency domain for signal analysis. The system structure of this method is very flexible, the sensitivity is high, and the speed generally depends on the back-end data acquisition system.Each of these demodulation methods has its own advantages and disadvantages. The actual application needs to be selected according to different requirements. For the measurement of vibration and dynamic strain of large structures such as bridges and rails, the application environment is relatively stable, and the measurement speed and range requirements may be relatively low. Multi-point monitoring with distributed sensing can improve efficiency and cost-effectiveness. Long-term use has high requirements for measurement accuracy, system stability, and actual data output. Scanning spectrum measurement methods or microwave spectrum analysis methods are optimal choices. However, for the instantaneous monitoring of the shock power of the explosion shock wave and the target load response, the single-point and high-speed acquisition methods are more suitable. Such transient measurements require high speed and range, the measurement time is short and signal can be collected first and then processed. Accordingly, the light intensity analysis method can be adopted. We can maximize the unique advantages of various demodulation methods and achieve efficient and accurate measurements only by selecting different high-speed demodulation methods according to different scenarios.

The spatial dispersion method is relatively mature. The Wasatch Cobra-S 800 spectrometer has a sampling rate of 250 kHz (Table 1). The measurement frequency of the systems based on the principle of time dispersion can up to 264 MHz (Table 1). In the scanning spectrometry measurement method, Tatsuya Yamaguchi et al. achieved a measurement frequency of 202.8 kHz driven by a conventional FDML laser with a scanning frequency of 50.7 kHz by processing the light source (Table 1). In the light intensity analysis method, Ding Z C et al. of Beijing Jiaotong University employed a cross-Sagnac loop as an edge filter to achieve a demodulation frequency of 200 kHz (Table 2). In the phase analysis method, Oton C et al. of the University of Florence achieved a demodulation speed of 100 kHz using an electro-optical modulator as a tunable retarder in the Sagnac loop (Table 3). In the microwave spectrum analysis method, Zhou Lei et al. used a cross-scan period to form a beat frequency, doubling the measurement frequency to 40 kHz. It has also been reported that the acquisition speed of this method can reach the megahertz level (Table 4).

This paper summarizes the main advantages and disadvantages of each method, the adaptation scenarios, and whether multi-grating measurement can be performed (Table 5); in addition, it sorts out some scenarios involving high-speed dynamic measurements and the corresponding high-speed demodulation methods (Fig.24), and looks forward to the future research direction of high-speed FBG demodulation methods. First, it is important to continue to improve the FBG signal demodulation speed, which requires further improvement of the speeds of signal acquisition and processing. The methods involve increasing the rate level of light sources, line array detectors, and other devices, and optimizing the demodulation algorithms. Second, it is important to improve the accuracy and demodulation range. The spectral range and resolution of the light source need to be increased. In addition, it is important to increase the FBG demodulation capacity. While improving the speed of spectral measurement, further expanding its spectral bandwidth is the basis for the development of quasi-distributed large-capacity FBG high-speed demodulation technology.

During the operation of an oil-immersed transformer, aging of the insulating cardboard, mechanical failure of the submersible pump, and action arc of the tap change introduce different particulate impurities into the transformer oil. If the particulate impurities suspend in the oil flow or adhere to the surface of the transformer windings and components, the safety of the transformer will be compromised. In recent years, laser-induced breakdown spectroscopy (LIBS) has been widely used for the detection of nonmetals and metals. There are also reports on the detection of particulate matter in transformer oil, and the filter-paper-assisted LIBS map-scanning method has proven to be effective for the quantitative analysis of particles. In this study, in addition to the target spectral lines, molecular bands and bremsstrahlung and recombination radiation in the plasma generated by laser ablation are present, and the resulting continuous background spectrum cannot be shielded during spectral analysis, which adversely affects the spectral intensity of the characteristic spectral lines of the target element. However, the ablation spots obtained by laser scanning the deposition area of the sample do not contain characteristic spectral lines of the target element, hence necessitating binary classifications to screen out the spectrally effective spots for quantitative analysis. Therefore, there is an urgent need for a set of data-processing algorithms for baseline correction, dimensionality reduction, and classification of the original data to meet the processing requirements of a large number of high-dimensional spectral data.

Based on the sparsity of characteristic spectral lines, this study investigates the application of the baseline estimation and denoising algorithm (BEADS) in LIBS spectral baseline correction and subtraction. The results of the dimensionality reduction using the target analysis line and principal component analysis (PCA) are compared. In addition, the binary classification effects of different machine learning algorithms (e.g., decision tree, support vector machine, K-nearest neighbor classification, and ensemble classifier) on laser ablation spots are studied, in which the ensemble classifier perfects the classifier model through different optimization methods. Finally, based on the above spectrum-processing algorithm, a quantitative analysis and calibration of Fe particle detection is completed.

Spectral analysis reveals that the characteristic spectral lines such as Fe Ⅰ 360.89 nm, Fe Ⅰ 361.88 nm, Fe Ⅰ 363.15 nm, Fe Ⅰ 364.78 nm, and Fe Ⅰ 371.99 nm have excellent sparsity, whereas Fe Ⅰ 373.49 nm, Fe Ⅰ 373.71 nm, Fe Ⅰ 374.83 nm, Fe Ⅰ 374.95 nm, and Fe Ⅰ 376.38 nm exhibit relatively poor sparsity owing to interference with each other or overlap with the background spectrum. The experimental results show that, as shown in Fig. 3, the application of asymmetric penalty functions and convex optimization techniques is beneficial for reducing overfitting. When the parameters are adjusted to f_c=0.15, d=1, r=6, and λ=0.8, the accuracy of the baseline estimation is very high, and the residual value obtained by the baseline fitting is small. For the detection of Fe particles in transformer oil, the BEADS algorithm can excellently deduct the continuous background, so that the intensity of the Fe characteristic spectral lines can be accurately corrected. As shown in Fig. 5, in terms of the classification accuracy of ablation spots by the classifier, the spectral data after dimensionality reduction using the target analysis line is better than the original spectral data and the spectral data after PCA is performed. This demonstrates that the dimensionality reduction processing method using the target analysis line is scientific and reasonable. Based on the decision tree algorithm, the Bagged Trees ensemble classifier constructs multiple decision tree models through the extraction of different samples, thereby reducing the variance, optimizing the classification model, and improving classification accuracy. The spectral classification accuracy for data after dimensionality reduction is as high as 98.33%.

Based on the spectral processing method, the linear correlation coefficient between the spectral correction intensity and particle mass ratio is 0.9983, and the relative standard deviation of repeated experiments is small, which proves the scientificity and robustness of the method. The method can realize batch processing of a large number of ablation spot data generated by laser map scanning, which greatly improves data-processing efficiency while reducing errors introduced by manual processing. It provides convenience for LIBS automatic acquisition and data processing and lays a theoretical foundation for LIBS detection of particulate matter in transformer oil.

The research purpose and focus of this paper is to propose a new signal processing method that processes a photon echo heterodyne signal and can achieve a higher signal-to-noise ratio (SNR) intermediate frequency (IF) signal spectrum and signal time-frequency characteristics with an improved performance to improve the photon counting heterodyne radar speed measurement performance of variable-speed moving targets.

This study applies the adaptive sparse degree compression perception method to a variable-speed target heterodyne photon echo signal and solves the problem whereby the sparse degree of K can not be determined in advance. The reconstructed frequency spectrum has a relatively high SNR but error and some unfiltered noise are also found on the spectrum. Furthermore, according to the characteristics of concentrated and continuous Doppler spectrum components of the variable-speed target, density clustering is creatively applied to the denoising of the above-reconstructed spectrum, and the sparsity adaptive compression sensing and clustering algorithm are combined as a new heterodyne signal processing method of photon echoes. In this paper, the first part of the proposed signal processing method is to solve the frequency spectrum. This process is divided into four parts, namely compressed perception reconstruction of the IF spectrum, density clustering, denoising, and interpolation. First, the frequency spectrum is reconstructed by the sparsity adaptive matching pursuit (SAMP) algorithm according to the photon arrival time series, the IF signal spectrum is reconstructed, and compressed perception processing is performed. The lower amplitude of the spectrum component is assumed as noise, and only discrete signals whose amplitudes are significantly higher than noise are retained in the reconstructed IF spectrum. However, at this time, the reconstructed IF signal spectrum is not the final signal spectrum, and the noise signal with higher spectrum amplitude is still retained. Compared with the noise, the signal of the ideal spectrum of the variable-speed target must be continuously changing, so the reconstructed IF signal spectrum is processed by density clustering. The frequency component with the highest density can be obtained by classifying and sorting the spectrum of the reconstructed IF signal according to the density-based clustering, which can be determined as the frequency component of the IF signal. In addition, discrete points can also be obtained while clustering the reconstructed IF spectrum, which can be regarded as noise components and then denoised. However, because the compressed sensing algorithm has removed most of the noise components, and through signal processing such as clustering denoising, the obtained IF signal spectrum becomes a discrete spectrum, which is inconsistent with the continuous spectrum of the ideal variable-speed moving target. Therefore, the IF signal spectrum is interpolated and fitted to obtain the final IF signal spectrum. The second part of the proposed signal processing method is to obtain a time-frequency characteristic analysis method with improved performance. According to the principle of short-time Fourier transform to obtain the time-frequency characteristic of signals, we combine the sparsity adaptive compression sensing in this method with the time-frequency characteristic analysis method in this paper.

The simulation results show that the proposed method has excellent advantages. Compared with the traditional direct Fourier transform spectrum, the SNR is improved by up to 20 dB on average, and the average accuracy error is within 10% (Fig. 7). It is found that the smaller the signal Doppler broadening, the more obvious the advantage of this signal processing method. When the signal Doppler broadening is gradually increased or the spectrum is complex, the reconstruction effect and density denoising effect will be lower than expected value (Fig. 8). Meanwhile, the contrast of time-frequency maps obtained by the proposed signal processing method is much higher than that of the short-time Fourier transform, wavelet transform and Wigner-Ville distribution, which indicates that the readability of the time-frequency maps obtained using the proposed method is the best (Fig. 9). From the two evaluation indexes of contrast and information entropy, the performance of the proposed time-frequency characteristic analysis method is seen to be the best (Fig. 11). Finally, experiments are performed to verify that the confidence interval of the SNR improvement value of the signal processing method in this paper is [8.5 dB, 11.1 dB] with a confidence interval of 95%. Further, the confidence interval of the Doppler broadening accuracy error is [5.6%,7.4%] (Fig. 14), and the time-frequency analysis characteristics are consistent with the simulation results that were obtained (Fig. 15).

The results show that the proposed method is effective at improving the SNR and time-frequency distribution of signals to optimize the speed measurement performance of the photon counting heterodyne radar against variable-speed moving targets. This method proves the feasibility of extracting the motion information or fretting information of the next complex moving target. Therefore, further research is needed to understand the extraction method of motion information for complex moving targets.

For combustion-driven deuterium fluoride/hydrogen fluoride (DF/HF) chemical lasers, fluorine atoms produced in the combustion chamber are the source of the laser energy. The production efficiency of fluorine atoms in the combustion chamber directly determines the upper limit of the chemical efficiency of combustion-driven DF/HF chemical lasers. The atomic fluorine production efficiency limits several characteristic parameters of combustion-driven DF/HF chemical lasers, such as the amplification scale, volume efficiency, and weight efficiency. Therefore, it is necessary to investigate the combustion process of combustion-driven DF/HF chemical lasers thoroughly. In the past, most studies on HF/DF laser combustors were conducted theoretically using thermodynamic equilibrium methods rather than experimentally because of the extremely high temperature of the combustion production gases and the strong corrosivity of the combustion products F, F₂, etc. To date, only a few indirect experimental studies have been conducted on HF/DF laser combustors. These experimental studies considered the laser output power as the research object to investigate the working performance of the combustion chamber indirectly without direct observation of the combustion process. In this study, a small combustion chamber platform was built. The ultraviolet–visible and near-infrared spectra of the H₂/NF₃ combustion flame were directly observed by flame fluorescence spectroscopy. The combustion process of the H₂/NF₃ mixture in the combustion chamber was analyzed using spontaneous emission spectroscopy. The gas temperature in the combustion chamber was measured using the rotational structure strength distribution of the HF (v=2→v=0) band. The flame temperature distribution along the gas flow direction is provided in combination with an electric translation platform. The influence of flow ratio of oxidant NF₃ to fuel H₂ on the flame gas temperature distribution was examined.

In this study, a small combustion chamber test platform was built (Fig. 1). The flame shape and temperature distribution of the gas products of H₂/NF₃ combustion in the combustion chamber were measured and analyzed by fluorescence spectroscopy. The entire combustion chamber is composed of corrosion-resistant stainless steel. The injector panel adopts a three-gas-jet layout, with a row of H₂ fuel injection holes in the middle and injection along the horizontal direction. There are two rows of NF₃ oxidizer injection holes, one at the top and one at the bottom, and the injection direction is 55° from the horizontal direction. Quartz observation windows are set on both sides of the gas flow passage in the combustion chamber to observe the combustion flame, and a large amount of N₂ curtain gas is injected on both sides to protect the quartz observation window. A pressure tap to monitor the pressure of the combustion chamber and an electric spark device for igniting the H₂/NF₃ mixture are set on the upper side of the combustion chamber. The electric spark device is located 35 mm downstream from the injector panel. A reflective optical fiber collimator coated with a silver film is used to collect the luminescence of the combustion flame transmitted from the quartz viewing window. The optical fiber collimator is fixed on an electric translation platform that can move at a uniform speed to measure the distribution of the fluorescence spectrum of the combustion flame along the gas flow direction. The ultraviolet-visible and near-infrared spectra of the combustion flame are measured using a grating spectrometer, and the wavelength of the spectrometer is calibrated using a pen-shaped low-pressure mercury argon lamp. A mass-flow controller is used to control and measure the flow rate of each gas.

In the ultraviolet–visible spectral region, the luminescence of the H₂/NF₃ combustion flame mainly includes the radiative transitions of electronically excited molecules, such as N₂(B), NF(b), and NH(A) (Fig. 3). In the near-infrared spectral region, the spectrum of the H₂/NF₃ combustion flame is relatively pure and simple and is mainly composed of the first overtone vibrational rotational transition band (Δv=2) of HF(v) vibrationally excited molecules (Fig. 4). At extremely high temperatures in the combustion chamber, the spectral lines at the high rotation quantum (J) overlap significantly, showing an apparent non-Boltzmann equilibrium. With the increase in the flow rate of oxidant NF₃, the start and end positions of the flame move upstream, but the overall flame length changes little (Fig. 7). The gas temperature in the combustion chamber rises rapidly at the beginning upstream, immediately reaches the highest temperature point, and then decreases slowly and linearly (Fig. 8). The highest temperature point is where the flame brightness is the highest and the combustion reaction is the most intense. When the flow ratio of NF3 to H2 gas is small, the gas temperature in the combustion chamber decreases more gently. When the flow ratio of NF₃ to H₂ gas increases, the gas temperature in the combustion chamber decreases more sharply.

A fluorescence spectrum analysis and temperature distribution measurements of the combustion flame of H₂/NF₃ mixture were performed. The results show that the fluorescence spectrum in the ultraviolet-visible spectral region is mainly produced by the radiative transitions of electronically excited molecules, such as N₂(B), NF(b), and NH(A). The fluorescence spectrum in the near-infrared spectral region is mainly composed of the first overtone vibrational rotational transition band of HF(v) vibrationally excited molecules. The effect of flow ratio of oxidant NF₃ to fuel H₂ on the H₂/NF₃ flame length was investigated. The experimental results show that, under the experimental conditions of the three gas jets, the flame length decreases when the flow ratio of oxidant NF₃ to fuel H₂ increases and increases when the flow ratio of oxidant NF₃ to fuel H₂ decreases. The effect of flow ratio of oxidant NF₃ to fuel H₂ on the combustion flame temperature distribution was also investigated. When the flow ratio of oxidant NF₃ to fuel H₂ is small, the gas temperature in the combustion chamber decreases gently along the flow direction. When the flow ratio of oxidant NF₃ to fuel H₂ gradually increases, the gas temperature decreases sharply along the flow direction.