An EEMD-SVD method based on gray wolf optimization algorithm for lidar signal noise reduction

Atmospheric lidar is susceptible to light attenuation, sky background light and detector dark current during detection, which results in a lot of noise in the lidar return signal. In order to improve the SNR and extract useful signals, this paper proposes a new joint denoising method EEMD-GWO-SVD, which includes empirical mode decomposition (EEMD), grey wolf optimization (GWO) and singular value decomposition (SVD). Firstly, the grey wolf optimization algorithm was used to optimize two parameters of EEMD algorithm according to moderate values: the standard deviation Nstd of adding Gaussian white noise to the signal and the number NE of adding Gaussian white noise. Secondly, the mode components obtained by EEMD-GWO decomposition are screened and reconstructed according to the correlation coefficient method. Finally, the SVD algorithm with strong noise reduction ability was used to further remove the noise in the reconstructed signal, and the lidar return signal with high SNR was obtained. In order to verify the effectiveness of the proposed method, the proposed method was compared with empirical mode decomposition (EMD), complete ensemble empirical modal decomposition (CEEMDAN), wavelet packet decomposition and EEMD-SVD-lifting wavelet transform (EEMD-SVD-LWT). The results show that the noise reduction effect of the proposed method was better than that of the other four methods. This method can eliminate the complex noise in the lidar return signal while retaining all the details of the signal. In fact, the denoised signal is not distorted, the waveform is smooth, the far-field noise interference can be suppressed and the denoised signal is closer to the real signal with higher accuracy, which indicates the feasibility and practicability of the proposed method.