Fault detection method for adaptive stochastic resonance bearing

Abstract

The invention discloses a fault detection method for an adaptive stochastic resonance bearing. The fault detection method for the adaptive stochastic resonance bearing comprises the following steps: firstly, preprocessing a fault signal, so that the fault signal meets a small-parameter stochastic resonance condition; secondly, according to a known condition, setting optimization range of an initial parameter and a bistable system parameter of an artificial fish swarm algorithm; thirdly, optimizing the value of the bistable system parameter by using the artificial fish swarm algorithm with a signal-to-noise ratio as an effect evaluating function; finally, adaptively finding the value of the bistable system parameter corresponding to the maximum signal-to-noise ratio, generating stochastic resonance and enhancing a resonance effect. The parameter optimization is performed by using artificial fish swarm algorithm, so that adaptive stochastic resonance is achieved; artificial intelligence and adaptive control are combined together, so that the application restriction of the stochastic resonance caused by difficult selection or incorrect selection of the system parameter is overcome and an early failure of the bearing can be effectively diagnosed.

Description

technical field [0001] The invention relates to a fault diagnosis method, in particular to a fault signal detection method used in bearing fault diagnosis. Background technique [0002] Rolling bearings are one of the most widely used mechanical parts, but also one of the most easily damaged components in mechanical equipment. According to statistics, in rotating machinery using rolling bearings, there are about Most of the mechanical failures are caused by bearings. If the early failure of the bearing is not diagnosed in time, it will cause serious failure of the machinery and equipment, resulting in huge economic losses. Therefore, diagnosing the early fault characteristics of bearings has great practical significance to avoid serious faults and ensure the normal operation of mechanical equipment. However, the features of early faults are very weak, and it is extremely challenging to extract the weak features of early faults. Most of the existing weak feature extracti...

AI Technical Summary

Problems solved by technology

However, the features of early faults are very weak, and it is very challenging to extract the weak features of early faults

Most of the existing weak feature extraction methods detect fault features from the perspective of eliminating noise, but for weak feature signals heavily polluted by noise, noise reduction can also weaken the weak feature signals while reducing noise, and the signal-to-noise ratio of the measurement signal still too low to detect

Images