Rolling bearing fault intelligent diagnosis method and system

Abstract

The invention provides a rolling bearing fault intelligent diagnosis method and a system. The rolling bearing fault intelligent diagnosis method comprises the following steps that a test signal and aan actual condition signal of bearing vibration are collected; 0-1 standardization processing is carried out on the test signal and the actual condition signal of bearing vibration, and a training setand a test set are divided respectively; an LSTM-softmax network is constructed, and the network is trained by using the test signal to obtain a source network suitable for a test; a dynamic time warping (DTW) distance between the test signal and the actual condition signal is calculated; and a target network is constructed on the basis of migrating all or part of the structure and parameters ofthe source network according to the DTW value, a new network is trained by using the actual condition signal, and finally the target network suitable for an actual working condition is obtained. According to the rolling bearing fault intelligent diagnosis method, the features of the vibration signals can be accurately extracted, and the problem of inaccurate feature extraction due to manual feature selection is avoided; and the problems that full-life-cycle data cannot be obtained under the actual working conditions, available data are few, and an effective diagnosis model is difficult to establish are solved, and the accuracy of fault diagnosis in actual production is improved.

Description

technical field [0001] The invention belongs to the field of fault diagnosis of mechanical equipment, and in particular relates to an intelligent fault diagnosis method and system for rolling bearings based on transfer learning and deep learning. Background technique [0002] Rolling bearings are an important part of mechanical equipment. A large proportion of mechanical equipment failures are bearing failures. If damaged bearings cannot be found and replaced in time, it will often affect the accuracy of the equipment, damage other components, and even cause serious production problems. ACCIDENT. Bearing damage is often caused by long-term work wear. Faults can be divided into: outer ring faults, inner ring faults, cage faults and ball faults. The frequency, amplitude and phase of bearing vibration signals are different for different faults. . Diagnosis of bearing faults based on vibration signals is currently the mainstream method. However, due to the nonlinear and non-st...

AI Technical Summary

Problems solved by technology

In view of the lack of available training data under actual working conditions, the difficulty of feature extraction, and the difficulty of accurately diagnosing bearing faults, the Long Short Time Memory (LSTM) network and the normalized exponential (softmax) regression model are used to extract and diagnose vibration signals. , using the idea of ​​transfer learning to migrate all or part of the structure and parameters of the source network, rebuild the network, and use live signal training to obtain a target network suitable for actual working conditions, so as to improve the accuracy of fault diagnosis in actual production and realizability

Images