中频雷达风速反演滤波降噪方法研究

Research on filter noise reduction method for wind speed retrieval of medium-frequency radar

摘要:

中频(medium frequency，MF)雷达是中高层大气风场探测的重要手段之一，广泛应用于风速反演和空间天气预测。雷达系统中广泛存在着噪声，其通过依次影响接收信号、相关函数、相关参数，最终使得MF雷达风速估计时存在偏差。为提高风速反演精度，提出一种高斯加权移动平均滤波方法(Gaussian weighted moving average filtering method，GW-MAFM)，对雷达接收信号进行降噪处理，改善风速反演的准确性。基于GW-MAFM反演风速误差小于MF雷达常用的拟合降噪方法，该算法在0 dB强噪声下的最大精度提升超过18 m/s 。与拟合降噪和中值滤波方法相比，GW-MAFM在精度和计算效率上均表现优异，有效扩展了MF雷达在强噪声环境中的风场反演的有效性。

Abstract:

Medium frequency (MF) radar is one of the important means for detecting wind field in the middle and upper atmosphere, widely used in wind speed retrieval and space weather prediction. There is widespread noise in the radar system, which affects the received signal, correlation function and correlation parameters in turn, and finally causes the deviation in the wind speed estimation of MF radar. To improve the accuracy of wind speed retrieval, a Gaussian weighted moving average filtering method (GW-MAFM) was proposed to reduce the noise of radar received signals and to improve the accuracy of wind speed retrieval. The error of wind speed retrieval based on GW-MAFM was smaller than that of MF radar, and the maximum accuracy of the algorithm was improved by more than 18 m/s at 0 dB strong noise. Compared with the fitting denoising and median filtering methods, GW-MAFM has excellent performance in terms of both accuracy and computational efficiency, effectively extending the effectiveness of MF radar wind field retrieval in strong noise environment.