Coherent Doppler wind lidar (CDWL) has emerged as an effective tool for analyzing wind velocity distributions. It utilizes the peak frequency of the signal spectrum to determine wind velocity. However, accurate identification of the spectrum peak in low signal-to-noise ratio (SNR) environments is complicated by noise pollution. Enhancing CDWL performance involves correcting the spectrum in these challenging areas. Existing probability-constraint-based methods (PCBMs) require empirical parameter settings, limiting their adaptability across different Doppler wind lidar environments. This paper proposes and demonstrates a probability-constraint-based method based on the honey badger algorithm (PCBM-HBA). The gate moving average method (GMAM) based on the spectrum obtains the reference wind velocity as a constraint. The correlation coefficient between the inverted wind velocity value of PCBM and the reference wind velocity is used as the negative value of the fitness function to obtain the optimal parameter σ. Simulation results based on the American Standard Atmosphere Model show that PCBM-HBA can measure wind fields in areas with low SNRs, and the maximum detection range improves from 3.8 to 5.4 km. During the inversion of the measured signal, the PCBM-HBA improves the inversion results of wind velocity under different pulse conditions, and the inversion results of the PCBM-HBA with 50 accumulated pulses are better than those of the traditional method with 150 accumulated pulses, which enhances the applicability of the PCBM and improves the performance of the system.
Read full abstract