Time-Frequency Characteristic Analysis And Identification Research For Lightning Rapid Electric Field Signals

Lightning is an atmospheric discharge phenomenon that is catastrophic.It can cause forest fires,damage buildings,damage communications equipment,and even threaten the lives of people and animals.Therefore,studying the occurrence and development of lightning,and the identification of different discharge categories have a certain guiding role for lightning warning and protection work.In this paper,four aspects of preprocessing,time-frequency analysis,feature extraction and classification identification of lightning fast electric field signals are studied.The purpose of this paper is to analyze the development law of lightning electric field by using modern signal processing technology,and to study the automatic identification based on the characteristics of lightning electric field.Lightning warning and protection provide the basis.The main work of this paper has the following three points:1.Aiming at the mixed noise in lightning signal,the empirical mode decomposition(EMD)and synchronous compression wavelet transform(SST)are applied to the denoising of lightning fast electric field signals,and a lightning electric field based on EMD and SST algorithm is proposed.Signal denoising method.At the same time,the method is applied to the denoising process of the simulated signal and the natural lightning signal respectively,and the denoising effect is compared with the wavelet threshold method,the EMD method alone,and the SST method alone.Experiments show that the denoising effect of this method is better.Compared with the other three methods,this method preserves the details of the lightning signal while removing noise.This is important for the study of lightning discharges such as backlash.The meaning.2.A new time-frequency analysis technique,Frequency Slice Wavelet Transform(FSWT),is applied to the time-frequency analysis of lightning signals.By introducing the frequency slicing function and the scale factor,FSWT can convert the one-dimensional time series of the lightning signal to the two-dimensional time-frequency plane,and simultaneously analyze the variation of the lightning signal energy with frequency and time in the time domain and the frequency domain.At the same time,the FSWT can take into account the time-frequency resolution of the low frequency band and the high frequency band and the time-frequency resolution of any control signal,and can extract and reconstruct the lightning signal in any frequency band,and provide a basis for the feature extraction of the lightning signal.3.The energy characteristics of the lightning signal are extracted by wavelet decomposition,empirical mode decomposition and frequency slice wavelet transform respectively,and they are sent as feature vectors to the support vector machine for automatic identification of ground flash and cloud flash.The experimental results show that the three methods have obtained a good recognition rate,and the recognition of the energy features extracted by the frequency slice wavelet transform has the highest recognition rate.