Bearing fault diagnosis method and system based on adaptive anti-noise neural network

Abstract

The invention discloses a bearing fault diagnosis method and system based on an adaptive anti-noise neural network, and the method comprises the steps: obtaining a data set, and dividing the data setinto a training data set and a testing data set according to a preset proportion; a neural network model is constructed, wherein the neural network model comprises an input layer, two convolution layers, two gating recursion unit layers, an attention mechanism layer and a deep neural network layer; training the constructed neural network model based on training data in a noiseless training data set to obtain a trained neural network model; and adding the additive white Gaussian noise into the test data in the test data set, and inputting the additive white Gaussian noise into the trained neural network model to obtain a bearing fault diagnosis result. Features can be automatically extracted from original signals, manual feature selection and denoising are not needed, different fault typesand fault severity degrees can be distinguished, and the method can adapt to different noise environments.

Description

technical field [0001] The invention relates to the technical field of bearing faults, in particular to a bearing fault diagnosis method and system based on an adaptive anti-noise neural network. Background technique [0002] Fault diagnosis of mechanical equipment has been widely concerned in modern industry, and the failure of mechanical equipment will cause economic losses and casualties. Rolling bearings are an important object of mechanical equipment fault diagnosis, especially in rotating mechanical equipment, rolling bearing faults account for a large proportion of faults. Fault diagnosis of rolling bearings has been extensively studied in the past decades. Data-driven methods are commonly used in bearing fault diagnosis, and are mainly divided into signal analysis-based diagnostic methods and deep learning-based diagnostic methods. [0003] Bearing fault diagnosis methods based on signal analysis mainly include: Fourier analysis method, wavelet transform method, ce...

AI Technical Summary

Problems solved by technology

However, in the early experiments of CNNs or RNNs for bearing fault diagnosis, complex manual operations were required, so the fault diagnosis of bearings in complex noise environments is still a challenging task, especially in the case of high noise. Adaptability to different noise levels still needs improvement

Images