A Signal Envelope Demodulation Method Based on Sliding Window Filter

Abstract

The invention discloses a signal envelope demodulation method based on sliding window filtering. Firstly, the signal is preprocessed, and the DC component in the signal is removed; ; Select the length of the sliding window, and according to the length of the sliding window, extend the boundary of the instantaneous energy sequence to meet the requirements of the sliding window filtering of the endpoint data. Then, the sliding window filter is performed on the extended sequence to obtain the envelope of the signal and realize envelope demodulation. This method is different from the traditional envelope demodulation method. The signal envelope demodulation method uses a sliding window to perform mean filtering on the instantaneous energy of the signal to extract the envelope information of the signal. It does not need to design a low-pass filter and has a high operating efficiency.

Description

technical field [0001] The invention belongs to the technical field of signal processing and feature extraction, and relates to a signal envelope demodulation method based on sliding window filtering. Background technique [0002] The phenomenon of amplitude modulation usually exists in dynamic signals of rotating machinery, such as bearings, gearboxes. When the rotating machinery fails, amplitude modulation phenomenon will occur. Therefore, extracting amplitude modulation features from dynamic signals can provide effective information for equipment monitoring. At present, commonly used methods include square demodulation, Hilbert envelope demodulation, Teager energy operator demodulation [1] and other methods. Square demodulation is a generalized detection filtering method. When square demodulation is implemented, the amplitude of the analysis signal is firstly squared, and then the squared signal is low-pass filtered to obtain the envelope of the signal to achieve envelop...

AI Technical Summary

Problems solved by technology

However, the square demodulation method needs to perform filtering when obtaining the envelope. In addition, there will be frequency mixing on the envelope spectrum; when the Hilbert envelope is demodulated, the analysis signal needs to meet certain conditions. In addition, the transformed Windowing processing is required during calculation, which will bring errors to the results, especially in the case of non-full-period sampling, this transformation will appear edge flying wing phenomenon

Images