Electromagnetic Signal Preprocessing and Feature Extraction

Abstract

The electromagnetic wave can lead to electrical equipment failure, which is really a serious problem to electrical power unit, which not only destroy the circuits but also may lead to fire. Therefore, This research focused on electromagnetic signal preprocessing and feature extraction based on wavelet transform. After building the model of electromagnetic waveform, the collecting signals of electrical spark affected by noise signals were processed, wavelet transform was used to filter the noise interference so that the features of electromagnetic waveform can be extracted. The result shows that the preprocessing of electromagnetic signal can achieve accurate feature parameter information and after extracting the features, which provide convenience to the further processing of electromagnetic signals.