The wind information of the middle and upper atmosphere is very important to study the coupling of the upper and lower atmosphere and energy, to ensure the smooth development of spacecraft space activities, and to carry out medium and long-term meteorological predictions. The doppler asymmetric spatial heterodyne wind measurement technology is a technique based on the Fourier transform of the interferogram to realize the detection of the doppler frequency shift of the wind. Doppler asymmetric spatial heterodyne is a new passive wind detection technology. For the interferometer, the processing and assembly errors of optical components and structural components, and the aberration of the optical system will distort the interference image. Introducing inversion error in the process of wind speed measurement. The current research on DASH interferogram distortion has not analyzed the influence of interferogram distortion on the accuracy of Doppler phase inversion and lacks the quantitative theoretical basis for the design, processing, and adjustment of Doppler asymmetric spatial heterodyne. In this paper, we analyzed the sources of different distortion in Doppler asymmetric spatial heterodyne. Then by adding different types and sizes of interferogram distortions to the interferograms of the red and green oxygen emission line, the simulation compares the difference between the distorted interferogram’s and the ideal interferogram’s Doppler phase. By adding optical distortion, local bending, slanted fringes and frequency changed these four different forms of interferogram distortion, we got the influence of distortion on the accuracy of Doppler phase inversion. The results show that the Doppler phase error will increase with the increase of the target wind field and interferogram distortion. The Doppler phase error of optical distortion is also will increase with the increase of the target wind field but will fluctuate increase with the increase of interferogram distortion. Among these four different forms of interferogram distortion, the local bending of fringes has the greatest influence on Doppler phase inversion. The phase error increases by 0.113‰ for each additional pixel of the local bending. But the maximum phase error is only 0.03‰ under the condition of 2% distortion. To further explore the influence of local bending sizes and location, we simulate various interferograms with local bending of different sizes and locations. The result shows that the Doppler phase error fluctuation decreases and gradually converges when the size increases. And the phase error fluctuates with the change of position. The fluctuation amount in the first half is small, and the fluctuation in the second half increases gradually. The phase error generated by the same bending at the sampling center is larger than that at the sampling edge. Therefore, attention should be paid to the small distortion in the sampling center area, and if necessary, interferogram correction should be performed to reduce the phase error. The simulation of errors caused by local bending on systems with different fringe frequencies shows that the same amount of bending will have a greater impact on systems with high fringe frequencies. In addition, interferogram with a low signal-to-noise ratio usually uses multiple rows of pixels of the interferogram to reduce uncertainty of phase. Local modulation is reduced when multiple rows of pixels of the distorted interferogram are merged. In order to find out the actual impact of the distorted interferogram in multiple rows of pixels of interferogram, we simulate different interferograms with local bending of different local bending max offset, in different signal-to-noise ratio and modulation. The result shows that even local modulation is reduced when multiple rows of pixels of the distorted interferogram are merged, but the phase uncertainty of the interferogram will not increase. Therefore, even if the interferogram has defects, multiple rows of pixels can be merged to increase the signal-to-noise ratio and reduce the phase uncertainty. This article may provide a quantitative theoretical reference for the design, processing, and adjustment of the Doppler asymmetric spatial heterodyne.