Rolling bearing fault prognostics method and system based on logistic regression and J divergence

Abstract

The invention discloses a rolling bearing fault prognostics method based on logistic regression and J divergence. The method comprises the following steps: fault sensing data at different fault positions in the case of bearing operation and normal sensing data for bearing operation in a normal state are acquired, preprocessing and feature extraction are carried out on the fault sensing data and the normal sensing data respectively, and fault position feature samples and normal state feature samples are built; through the well-built fault position samples and the normal state samples, a logistic regression model is trained, logic model parameters are obtained, and a logistic regression model is built. Targeted fault diagnosis is carried out on the bearing according to the health decline degree of the bearing, parameters such as a training sample type, a feature value type and a health threshold are changed according to different working conditions and different objects, the trained model can be adjusted, and the method and the system have the advantages of strong real-time performance, high data processing precision, good robustness of a core algorithm, high state evaluation accurate, high diagnosis result accuracy and the like.

Description

technical field [0001] The invention relates to the technical field of mechanical product quality reliability evaluation and fault diagnosis, in particular to a method and system for fault prediction of a rolling bearing based on logistic regression and J divergence. Background technique [0002] Rolling bearing is an important support of rotating machinery, and its real-time operating state directly determines the reliability of mechanical equipment. Therefore, it is very important to study the online fault diagnosis of rolling bearing. Real-time evaluation of the current operating state of rolling bearing, locating the fault location and quantifying the severity of the fault, to The maintenance and design of mechanical equipment has extremely important guiding significance. This patent implements an online fault prediction method based on logistic regression and J divergence for rolling bearings, which can evaluate the running state of the rolling bearing in real time and ...

AI Technical Summary

Problems solved by technology

[0004] The present invention aims at the disadvantage of being unable to accurately evaluate the running state of the bearing in the prior art, and provides a rolling bearing fault pre-diagnosis method and system based on logistic regression and J-divergence

Images