Research On Fault Diagnosis Method For Subway Gear Bax Based On Variational Mode Decomposition And Stochastic Resonance

With the rapid development of rail transit,the bearing as key components of subway vehicle gearbox,its safe and smooth operation directly influences the safety of life and property of the passengers.It is of important practical and economic value to develop and study the fault diagnosis algorithm of subway vehicle gearbox bearing.As a typical rotating equipment,gearbox bearings often work under harsh conditions.And its vibration signal has non-stationarity and is often drowned by strong noise,lead to the vibration signal collected cannot realize fault diagnosis through the traditional time-frequency analysis method.In practice,it is difficult to extract fault feature information effectively because the vibration signal of the parts of the subway gear box is polluted by strong noise.Based on the vibration analysis and modern fault diagnosis method,the author has carried out a series of research work on the basis of vibration analysis:(1)For parts of the gear box have very low signal to noise ratio and complex component,and in order to overcome the problem of modal aliasing and owe envelope problems existing in traditional mode decomposition methods,a method for denoising of adaptive variational mode decomposition based on particle swarm was proposed.By comparing with empirical mode decomposition method,this new proposed method can effectively overcome the modal aliasing and improve the signal-to-noise ratio.(2)To overcome the adiabatic approximation theory limits on stochastic resonance method,through the research analysis bistable stochastic resonance method,a new self-adaptive step-changed stochastic resonance(SCSR)based on particle swarm optimization was proposed.By using the fast global optimization feature of particle swarm algorithm,this proposed method can optimize the structure parameters and calculation step of SCSR to adaptive achieve the optimal output effect.The method is applied to simulation data and actual bearing data,and the results show that this method can realize fault diagnosis more accurately than traditional stochastic resonance method.(3)For the bearing and gear signals collected in actual condition is polluted by strong noise,a integrated method of bearing fault diagnosis combining VMD with SCSR based on PSO is proposed.Firstly,this method decomposes the original signal with VMD and obtains several Band Limited Intrinsic Mode Function(BIMF)and the BIMFs were reconstructed by using Kurtosis criterion;The structure parameters and calculation step of SCSR are adaptive determined by using particle swarm algorithm and the reconstruction signal was analyzed by SCSR.The numerical simulation experiment and the analysis of bearing and gear failure data show that this method can more effectively highlight the fault characteristic signal and verify the effectiveness of the method.