Rolling bearing fault diagnosis method based on blind signal separation and support vector machine

Abstract

The invention provides a rolling bearing fault diagnosis method based on blind signal separation and a support vector machine, belongs to the technical field of intelligent fault diagnosis of rotatingmachinery, and aims to improve the precision, efficiency and robustness of rolling bearing fault diagnosis, and the implementation steps are as follows: obtaining a training sample set and a test sample set; obtaining a multi-dimensional feature vector set corresponding to the training sample set; obtaining a support vector machine set; performing iterative training on the support vector machineset; defining an observation matrix and a separation matrix; performing blind signal separation on the observation matrix; and obtaining a fault diagnosis result of the rolling bearing. According to the invention, the self-adaptive selection nonlinear function and the iteration step length are introduced into the iteration of the separation matrix, the multi-dimensional feature vector composed ofthe amplitude domain parameter, the frequency domain index and the multi-scale entropy is adopted when the feature vector of the vibration signal is extracted, and the diagnosis precision, efficiencyand robustness are effectively improved in combination with the support vector machine.

Description

technical field [0001] The invention belongs to the technical field of intelligent fault diagnosis of rotating machinery, and relates to a rolling bearing fault diagnosis method, in particular to a blind signal separation method based on a nonlinear function and a step size that can be adaptively selected and a support vector machine that can identify according to features The rolling bearing fault online diagnosis method can be applied to judge the specific fault location of the rolling bearing. Background technique [0002] Rolling bearing is an important part in mechanical system, and it is also a part prone to failure. Common bearing faults are divided into outer ring faults, inner ring faults, rolling element faults, cage faults and composite faults according to fault types. Severe bearing faults will cause the failure of the overall mechanical system. Therefore, timely diagnosis and determination of the specific fault location of the bearing in the early stage of the ...

AI Technical Summary

Problems solved by technology

[0006] For example, Wang Hairui et al. published a patent application with the publication number CN111027259A in 2019, entitled "A Rolling Bearing Fault Detection Method Based on Dragonfly Algorithm Optimizing BP Neural Network", and disclosed a combination of wavelet packet and BP network to complete rolling bearing fault detection. The method of fault diagnosis, the method collects the vibration signal during the operation of the bearing through the sensor, decomposes the collected signal through the wavelet packet to obtain the feature vector, and inputs it into the trained BP neural network to complete the fault detection of the rolling bearing. The existing technology has solved the problem that the BP neural network is easy to fall into the local extremum. However, the feature extraction of the vibration signal by the wavelet packet method cannot fully obtain the characteristics of the signal, and the lack of features will lead to the problem of low diagnostic accuracy, and the optimization goal of the BP network is Based on empirical risk minimization, a large number of samples are required for training and the training results are unstable, which will lead to low diagnostic efficiency and poor robustness.

Images