Time Domain Feature Rapid Analysis For High Frequency Signal In Electrical Apparatus

High frequency signal components are often detected in on-line measurement of electrical apparatus,and high frequency signals contain abundant partial discharge information.At present,the on-line measurement system of electrical equipment has long interval of collecting data and weak analysis performance,so it is urgent to apply advanced signal processing methods to on-line measurement system,collect more data,extract more useful information from it,and gradually realize the purpose of learning in measurement.This paper first defines the signal collected by the sensor in the on-line measurement system as the high frequency signal,and then introduces the two level analysis model of the high frequency signal of the generator,that is to analyze the characteristic of the high frequency signal first,and then select the characteristic signal for the analysis of the partial discharge signal.Then,the historical measurement data of generators are summarized and classified,and the characteristics of high-frequency signal data are analyzed.A set of time-domain characteristic parameters of high frequency signals is obtained,and the characteristics of different kinds of pulses are described by using this set of characteristic quantities.Then,the high frequency signal feature analysis algorithm is designed.The algorithm includes extracting the characteristics of the waveform,analyzing the extraction results,and judging whether the waveform is stored.The algorithm simulates the steps of the human eye recognition signal,extracts the characteristic points of the information of the reaction waveform from the original signal,and then statistics the characteristics of the feature points,and obtains the fluctuation level of the current signal.In this paper,we introduce two methods of amplitude division window and phase division window to judge whether the waveform changes or not,and design the relevant storage strategy.Finally,the high frequency signal feature analysis algorithm is tested.First,simulation tests on the MATLAB platform verify the accuracy of the algorithm.Then the algorithm is embedded in the partial discharge signal analysis system and tested in laboratory environment to verify the accuracy,rapidity and stability of the algorithm.Finally,the algorithm is tested in the field environment,and the effect of the algorithm is evaluated according to the operation.