Antifriction bearing performance degradation assessment method based on empirical mode decomposition and logistic regression

Abstract

The invention discloses an antifriction bearing performance degradation assessment method based on empirical mode decomposition and logistic regression, comprising the step of: first, extracting the intrinsic mode function energy of bearing vibration signals as a characteristic vector; second, building a logistic regression model by utilizing the characteristic vectors of a bearing normal state and a failure state to obtain regression parameters; and finally, calculating the assessment index of a bearing signal life cycle, and thereby assessing antifriction bearing performance degradation conditions. According to assessment results, the method can timely discover early faults, well describe each phase of bearing performance degradation, and facilitate early detection of bearing initial faults and failure critical points so as to perform condition-based maintenance.

Description

technical field [0001] The invention relates to a rolling bearing performance degradation evaluation method based on empirical mode decomposition and logistic regression, and belongs to the technical field of mechanical product quality reliability evaluation and fault diagnosis. Background technique [0002] With the advancement of science and technology and the development of industry, in order to improve production efficiency and reduce production costs, industrial machinery and equipment are increasingly developing in the direction of large-scale, high-speed, systematization and automation. The functions of key equipment structures are becoming more and more complex, the working environment is more harsh, and the potential for failure during long-term operation also increases accordingly. Once a key component of the equipment fails, it may cause the entire equipment to crash and shut down, affecting the entire production process and causing huge economic losses. Therefor...

AI Technical Summary

Problems solved by technology

However, no matter the bearing performance degradation evaluation method based on neural network or hidden Markov model, the feature extraction method has not fundamentally overcome the shortcomings of Fourier transform. adaptive

In addition, when a fault occurs, the bearing vibration signal exhibits complexity and non-stationarity, leading to limitations in conventional feature extraction methods

Images