Fault Feature Extraction Method of Rolling Bearing Based on Hierarchical Sparse Coding

Abstract

The invention discloses a hierarchical sparse encoding-based rolling bearing fault feature extraction method. The method comprises the following steps of: carrying out high-frequency sampling on a bearing vibration signal, and intercepting a signal in a period of time as an original time domain signal; constructing a fixed dictionary sparse encoding module, setting related parameters, carrying outsparse encoding by using the fixed dictionary sparse encoding module so as to obtain a sparse representation coefficient matrix, and carrying out multiplication through a fixed dictionary and the sparse coefficient matrix so as to obtain harmonic interference components unrelated to a fault signal; filtering the harmonic interference components from the original time domain signal so as to obtainan input time domain signal; constructing a K-SVD-based dictionary learning model, taking the input time domain signal as an input of the model, and carrying out feature extraction by using the modelaccording to the set parameters so as to obtain fault-related features; and carrying out Hilbert transformation and fast Fourier transformation on the fault-related features in sequence so as to obtain a fault-related feature envelope spectrum, and outputting a diagnosis result. The method has the effects of improving the signal to noise ratio and filtering unrelated signals.

Description

technical field [0001] The invention relates to the technical field of rolling bearing fault diagnosis, in particular to a rolling bearing fault feature extraction method based on layered sparse coding. Background technique [0002] Rolling bearings are the most common and common rotating parts in modern industry, and their health status has an extremely important impact on the normal operation of the whole equipment. It is of great significance to effectively detect and diagnose early bearing failures. Rolling bearing vibration signals contain a lot of bearing health status information, but these information are usually interfered by some rotating parts (gears, shafts) and background noise. Therefore, how to extract effective feature components from these signals to accurately determine whether the bearing is faulty and The fault level line is particularly important. [0003] In recent years, a large number of signal processing methods have been proposed for noise reductio...

AI Technical Summary

Problems solved by technology

[0003] In recent years, a large number of signal processing methods have been proposed for noise reduction and fault feature extraction of rolling bearing vibration signals. The deep learning method based on dictionary learning (K-SVD) has also attracted extensive attention and research in recent years. The adaptive learning of the signal extracts the effective information in the signal and greatly improves the signal-to-noise ratio of the signal. However, because the fault signal cannot be distinguished from the irrelevant signal, the learned signal often contains some harmonics that are not related to the fault. components, such as periodic fluctuations caused by the rotating shaft, etc.

Images