非相干多普勒测风激光雷达信号去噪方法研究
Research on Signal Denoising Method of Non-Coherent Doppler Wind Lidar
蒋雨辰 1衣文索 1李东旭 1王鑫睿 1韩冬子1
作者信息
- 1. 长春理工大学光电工程学院,吉林长春 130022
- 折叠
摘要
为提高多普勒测风激光雷达探测精度,针对回波信号的背景噪声干扰大的问题,进行去噪方法研究.搭建基于马赫曾德鉴频器的非相干激光雷达系统,实验对比FIR滤波器傅里叶分析、滑动平均滤波和软、硬阈值法小波去噪.结果表明,半软阈值的去噪算法处理信号后信噪比较其他方法提升更大,均方根误差更小,时域和频域结合更紧密.因此,将半软阈值的小波去噪算法用于需要测量频移来获取风速的激光雷达上更有优势.
Abstract
To enhance the detection precision of Doppler wind lidar,this study addresses the challenge posed by significant background noise interference in echo signals.A Mahzender frequency detector-based incoherent lidar system is developed,and a comparative analysis is conducted on various denoising methods.The study evaluates the effectiveness of Fourier analysis,FIR filtering,moving average filtering,and wavelet denoising with both soft and hard thresholds.Results indicate that the semi-soft threshold denoising algorithm outperforms other methods,yielding reduced signal noise and smaller root mean square error.Additionally,it effectively integrates time and frequency domains.Thus,employing the wavelet denoising algorithm with semi-soft thresholds is advantageous for lidar applications requiring precise frequency shift measurements for wind speed determination.
关键词
多普勒测风/激光雷达/小波分析/信号处理Key words
Doppler wind measurement/laser radar/wavelet analysis/signal processing引用本文复制引用
出版年
2024
NETL
NSTL
万方数据