• Journal of Radiation Research and Radiation Processing
  • Vol. 42, Issue 1, 010601 (2024)
Zhaojin LUO1, Chengfeng LIU1, Wenbao JIA1、2, Qing SHAN1, Chao SHI1, Jiandong ZHANG1, Daqian HEI3, Xiaojun ZHANG4, and Yongsheng LING1、2、*
Author Affiliations
  • 1Institute of Nuclear Analysis Technology, Nanjing University of Aeronautics and Astronautics, Nanjing 211106, China
  • 2Jiangsu University Collaborative Innovation Center for Radiation Medicine, Suzhou 215031, China
  • 3College of Nuclear Science and Technology, Lanzhou University, Nanjing 211106, China
  • 4Suzhou Guanrui Information Technology Co., Ltd., Suzhou 215008, China
  • show less
    DOI: 10.11889/j.1000-3436.2023-0093 Cite this Article
    Zhaojin LUO, Chengfeng LIU, Wenbao JIA, Qing SHAN, Chao SHI, Jiandong ZHANG, Daqian HEI, Xiaojun ZHANG, Yongsheng LING. Complete coverage path planning of nuclear radiation field using bio-inspired neural network[J]. Journal of Radiation Research and Radiation Processing, 2024, 42(1): 010601 Copy Citation Text show less

    Abstract

    Path planning for the complete coverage of nuclear radiation fields is necessary to ensure the radiation safety of regional operators in radiation environments. Based on a bio-inspired neural network algorithm, a complete coverage path-planning algorithm for the optimal control of the radiation dose is proposed. First, part of the terrain of the Fukushima nuclear power plant and the Monte Carlo particle transport program were used to construct the obstacle distribution and radiation dose field in a simulated nuclear radiation field. Subsequently, the Python programming language was used to conduct algorithm simulation experiments. Each grid of the simulated nuclear radiation field was defined as a neuron, and a bio-inspired neural network was established. The grid dose rate and neuronal activity were combined to achieve optimal control of radiation dose in path planning, and single, four, and eight mobile units were used for simulation experiments. The results showed that the planning path of a single mobile unit can achieve 100% coverage and a 4% coverage repetition rate and can first cover the low-dose area and delay the coverage of the high-dose area to achieve optimal control of the process and cumulative doses. The algorithm is improved via a multiunit collaborative search to increase the time efficiency of complete coverage and decrease the cumulative dose of monomers. The coverage repetition rates of four-unit and eight-unit simulations were 5.72% and 6.29%, respectively. The complete coverage times of the one-unit, four-unit, and eight-unit simulations were 30, 9, and 4 min, respectively, and the time efficiency was doubled. The maximum cumulative doses of the monomers for the one-unit, four-unit, and eight-unit simulations were 4.11×10-3, 1.28×10-3, and 0.85×10-3 mSv, respectively, which also decreased significantly. The proposed algorithm can achieve complete coverage path planning of optimal control of the process dose and cumulative dose. Moreover, the algorithm can coordinate multiunit path planning and significantly decrease the cumulative dose of monomers, which is critical for radiation protection during regional operations in a radiation environment.
    Zhaojin LUO, Chengfeng LIU, Wenbao JIA, Qing SHAN, Chao SHI, Jiandong ZHANG, Daqian HEI, Xiaojun ZHANG, Yongsheng LING. Complete coverage path planning of nuclear radiation field using bio-inspired neural network[J]. Journal of Radiation Research and Radiation Processing, 2024, 42(1): 010601
    Download Citation