Variable working condition rolling bearing fault diagnosis method for optimizing theme correlation analysis

Abstract

The invention discloses a variable working condition rolling bearing fault diagnosis method for optimizing theme correlation analysis, and the method comprises the steps: firstly carrying out the feature extraction of rolling bearing vibration signals in a source field and a target field through wavelet transform, and constructing feature vectors to form a fault sample set; adjusting the value of a taste concentration judgment value in the fruit fly optimization algorithm and adopting a multi-population strategy to realize optimization of the main parameter total topic number and the common topic proportion with limited value range in the TCA algorithm; and finally, constructing an improved multi-population fruit fly optimization algorithm to optimize a topic correlation analysis IFOA-TCA fault recognition model, and realizing training of the vibration data of the rolling bearing on the fault recognizer under other working conditions and fault recognition of the trained fault recognizer on the target rolling bearing. According to the method, a good fault diagnosis result can be obtained under various working conditions, and the method has good engineering value and application prospects.

Description

technical field [0001] The invention relates to a rolling bearing fault diagnosis technology, in particular to a variable working condition rolling bearing fault diagnosis method for optimizing subject correlation analysis. Background technique [0002] As one of the key components of the mechanical transmission system, bearings play an important role in modern industrial production. However, during the operation of equipment, they may face uncertain working conditions and harsh environments, making bearings withstand unpredictable The instantaneous damage of the bearing will have a serious impact on the life of the bearing. Therefore, it is helpful to prevent major accidents from frequently carrying out fault diagnosis on the bearing. Facing the problem of rolling bearing fault diagnosis under variable working conditions, the traditional intelligent fault diagnosis method based on a large amount of data generally has the following defects: 1) The distribution of training da...

AI Technical Summary

Problems solved by technology

Facing the problem of rolling bearing fault diagnosis under variable working conditions, traditional intelligent fault diagnosis methods based on large amounts of data generally have the following defects: 1) The distribution of training data and test data is inconsistent, and cannot be directly used for fault diagnosis; 2) The target field of rolling bearing There are not enough fault samples to train a good fault classifier

However, transfer learning is still in the exploratory stage in the field of rolling bearing fault diagnosis, and the means used are limited. New transfer learning methods suitable for rolling bearing fault diagnosis with high fault recognition rate are worth exploring

Images