Research On Bearing Fault Diagnosis Based On Stochastic Resonance System

Rolling bearings are the main parts of mechanical equipment.Once rolling bearings fail,a series of safety problems will occur.Therefore,fault diagnosis and detection of rolling bearings have become the top priority of current industrial research.stochastic resonance(SR)can transfer energy of noisy signals to fault signals to reduce the impact of stochastic resonance on fault signals,and can effectively extract weak fault signals mixed with noisy environments.In this thesis,the basic theory of classical stochastic resonance is studied first and the transition phenomenon of overdamped Brownian particles in the system potential function is described.In order to solve the problem that the traditional stochastic resonance can only analyze weak signals,this thesis introduces the frequency shift and variable scale method to adjust the time scale,and uses the fourth-order Runge-Kutta equation to solve the system output signals and identify the bearing fault types.In addition,Quantum Particle Swarm Optimization(QPSO)was used to optimize the structural parameters of stochastic resonance system.The results show that the improved stochastic resonance bearing fault diagnosis method can solve the problem that the traditional stochastic resonance can only analyze weak signals.Then,a mathematical model of tristable stochastic resonance is proposed.The potential function of the tristable stochastic resonance system,Langevin equation under noise and non noise conditions are theoretically derived.The transition phenomenon of overdamped Brownian particles in the system under the joint action of noise signal and periodic signal is described.The correlation between potential well,barrier height and system structure parameters is analyzed.Then,a method of bearing fault diagnosis based on tristable stochastic resonance is constructed by combining tristable stochastic resonance with frequency shift scaling method and quantum particle swarm optimization algorithm.The results show that this method is effective for bearing fault feature extraction under strong noise,and provides a new idea for rolling bearing fault diagnosis.Finally,based on MATLAB GUI technology platform,the software design of improved stochastic resonance bearing fault diagnosis method and tristable stochastic resonance bearing fault diagnosis method is carried out,and a bearing fault diagnosis system software based on stochastic resonance system is built.The bearing data provided by the official website of Case Western Reserve University in the United States are used to verify the bearing fault diagnosis system based on the stochastic resonance system.The results show that the bearing fault diagnosis system based on stochastic resonance system can effectively extract the fault characteristic frequency in the bearing data,which lays a foundation for the intelligent diagnosis of rolling bearings(extraction and analysis of weak signals under strong noise).