Rolling bearing fault diagnosis method based on wavelet and improved PSO-RBF neural network

Abstract

The invention relates to a rolling bearing fault diagnosis method based on wavelet and improved PSO-RBF neural network. The method comprises the following steps: S1, reading an original vibration signal of a rolling bearing state of a rail train; S2, carrying out wavelet denoising on the original vibration signal to obtain a pure vibration signal; S3, carrying out decomposition and phase space reconstruction on the pure vibration signal by using a wavelet packet, and extracting a feature vector; S4, training an improved PSO-RBF neural network by taking the feature vector as input and the corresponding bearing state as output to obtain a neural network fault diagnosis model; and S5, for a to-be-diagnosed original vibration signal of the rolling bearing of the rail train, inputting the feature vector obtained after the steps S1-S3 into the neural network fault diagnosis model obtained in the step S4 to obtain a diagnosis result of the rolling bearing. Compared with the prior art, the method has the characteristics of integrating the advantages of wavelet transform and an artificial intelligence algorithm, and can carry out accurate fault diagnosis on the rail transit running gear.

Description

technical field [0001] The invention relates to the technical fields of artificial intelligence and rail transit fault diagnosis, in particular to a rolling bearing fault diagnosis method based on wavelet and improved PSO-RBF neural network. Background technique [0002] In recent years, the research on the fault diagnosis of rail train rolling bearings is very popular. For example, the Fourier transform is used to extract the characteristics of vibration signals. Since the fast Fourier transform (FFT) can realize the rapid transformation of the signal from the time domain to the frequency domain, it is based on Various spectrum analysis of FFT, such as refined spectrum analysis, maximum entropy spectrum analysis, holographic spectrum analysis, etc., have played a huge role in bearing condition monitoring and fault diagnosis; use historical data samples to train BP neural network or radial basis neural network Network (Radial Basis Function Neural Networks, RBF neural networ...

AI Technical Summary

Problems solved by technology

The convergence speed of BP neural network is slow, there are too many parameters to be optimized, and it is difficult to find the optimal parameters of RBF neural network structure

When using the principal component analysis method, when the sign of the factor loading of the principal component is positive or negative, the meaning of the comprehensive evaluation function is not clear

Images