搜索

x

基于小波分析对地面电场信号的去噪及闪电信息提取

Ground Electric Field Signal Denoising and Lightning Information Extraction Based on Wavelet Analysis

  • 摘要: 基于小波分析方法,通过对7种小波函数在误差控制和阈值方法选取上的对比,参考信噪比及均方误差等计算结果,选取了较为合理的Rigrsure阈值分析方法和sym5小波函数作为地面大气电场信号的去噪处理方法.该方法不仅能够较好地对电场信号波形平滑处理,降低电场波形中噪声信号的叠加度,还可保留闪电引起的电场的快速变化.在此基础上,利用差分计算方法实现了对闪电信号的自动识别,引入了总闪电频数误差及分时段闪电频数与人工分析结果的相关系数等2个参考量来选取差分计算方法中电场差分阈值,通过该方法可以实现闪电频数和极性等有关信息的提取.

     

    Abstract: In the report, based on the wavelet analysis method, the error control and selection of threshold method of even kinds of wavelet functions were compared, and the calculation results of SNR and mean square error were used as the reference, the more reasonable Rigrsure threshold analysis method and sym5 wavelet function were selected as the de-noising method for ground atmospheric electric field signal. The method not only smooth the waveform of electric field signal and reduce noise signal mixed into electric field waveform significantly, but also keep the characteristic of fast change of electric field caused by lightning flash. Based on it, the difference calculation method was used to realize the function of automatic recognition of lightning signals from the electric field waveform. Two reference values, the total lightning frequency error and the correlation coefficient between automatic recognition and manual analysis results in a continuous time range, were introduced to select the electric field difference threshold in the differential calculation method, which can extract the information about lightning frequency and lightning polarity.

     

/

返回文章
返回