Motor fault diagnosis method and system based on cavity convolution capsule network

Abstract

The invention relates to a motor fault diagnosis method and system based on a cavity convolution capsule network, and the method comprises the following steps: (1), obtaining a training sample with alabel, wherein the training sample comprises a motor vibration signal and a corresponding operation state, and the operation state comprises a normal state and a fault type in a fault state; (2) establishing a cavity convolution capsule network, and performing training by using the training sample; and (3) acquiring a to-be-diagnosed motor vibration signal, inputting the to-be-diagnosed motor vibration signal into the trained cavity convolution capsule network, and outputting the operation state of the motor. Compared with the prior art, the method has the advantages that the effective features of the motor signals can be automatically extracted, intelligent fault diagnosis is achieved, the diagnosis accuracy reaches 99% or above, the robustness and generalization ability are high, and theerror recognition rate is remarkably reduced.

Description

technical field [0001] The invention relates to a motor fault diagnosis method and system, in particular to a motor fault diagnosis method and system based on a hollow convolution capsule network. Background technique [0002] At present, there are mainly the following methods for motor fault diagnosis: [0003] Method 1: Motor fault diagnosis method based on signal processing. It preprocesses the collected signal to eliminate noise, reduce feature dimension and extract useful fault feature information. Signal processing methods mainly include Fourier transform, wavelet transform, wavelet packet transform and empirical mode decomposition methods. [0004] Method 2: Motor fault diagnosis method based on expert system. It uses the knowledge and reasoning methods of many human experts and scholars to deal with complex problems. Summarize the fault signal into a rule and establish an expert knowledge base. When a fault occurs, use the empirical analysis and reasoning in the...

AI Technical Summary

Problems solved by technology

[0007] Insufficiency of method 1: In the process of signal processing, not only a large amount of signal processing expertise is required to process and analyze the signal, but also technical personnel are required to have strong professional knowledge on the operating status and fault related background of the detected motor , coupled with the existence of a large number of unpredictable factors during the operation of the motor, the fault diagnosis process is more complicated, and the fault characteristic signal is not obvious and is non-linear and non-stationary, which increases the possibility of human judgment errors

[0008] Insufficiency of method 2: It is difficult to establish a relatively complete motor fault knowledge base based on the expert system diagnosis method. At the same time, the system does not have the ability of self-learning, and its robustness is poor.

[0009] Insufficiency of method 3: some traditional machine learning methods, such as support vector machine, naive Bayesian, decision tree and other methods, need to set the initial value of the parameters manually, the optimization process of the parameters is slow, and it is difficult to find the optimal parameters To match the optimal model, it is necessary to combine genetic algorithm, particle swarm optimization and other parameter optimization algorithms to assist in the optimization; for the BP neural network, it is easy to fall into the local minimum, and the deepened deep learning network model is prone to overfitting although its learning ability is improved. phenomenon, while increasing the complexity of the model

In the context of motor big data, for complex classification problems, the feature extraction and generalization capabilities of signal processing methods and traditional machine learning methods are subject to certain constraints, which can no longer meet the required requirements

Images