Automatic Fault Detection Method of Rotating Machinery Bearing Based on Complexity Analysis

Abstract

In order to improve the working condition stability of rotating machinery bearings, an automatic fault detection method of rotating machinery bearings based on complexity analysis is proposed. The parameters of rotating machinery bearings are detected by using discrete sensor detection method, and the fault characteristics of rotating machinery bearings are extracted by fault type parameter extraction method. The complexity of rotating machinery bearings is detected by wavelet transform and fuzzy logic analysis. Considering the load mutation, noise disturbance, control strategy and other factors of rotating machinery bearings, the fault diagnosis and feature analysis model of rotating machinery bearings is established. The clustering and feature detection of rotating machinery bearings’ fault types are realized by the method of discrete complexity fusion clustering. Based on adaptive feedback compensation and parameter estimation, the output diagnostic error is taken as the fault detection index parameter, and the feature estimation in normal state and fault state is realized by harmonic level detection, so as to realize automatic fault detection of rotating machinery bearings. The simulation results show that this method has high accuracy and small deviation in fault detection of rotating machinery bearings, which improves the ability of automatic fault diagnosis.