Study On Fault Diagnosis Of Gearbox Based On Particle Filter

The research is originated by National Natural Science Fund Project"complex gear fault diagnosis of early based on particle swarm optimization and filtering technique"(project NO: 50875247) .The gearbox is most commonly used the transmission device in the mechanical device, usually is in the long-term load to revolve, the error that because in the manufacture and assembly, as well as weary, aging and other effect existences, the gear box break down in the work, but the condition directly relates to the movement of entire equipment, therefore it has the significant significance to carries on the condition monitor and failure diagnosis.The particle filtering is one model-based state estimation technique solution non-Gauss misalignment stochastic system question. The particle filter is applied to fault diagnosis of gearbox can solve the non-Gaussian nonlinear problem. Using particle filter needs to know the system state space model. Establish time series ARMA model of the vibration acceleration signal, then the coefficients of this ARMA model as the particle filter coefficients of the state equation.ARMA model order is established by useing of FPE guidelines, then useing the least square method to estimate the parameters. The simulation results show that the particle filter in the signal denoising is feasible, using particle filter of the normal and fault conditions of the vibration acceleration signal noise reduction, comparative analysis of characteristics of the data values before and after noise reduction, noise reduction eigenvalues after the data better than the former.Study on fault diagnosis based on PF neural network,based on the algorithm of PF neural network optimization neural network model. Extracting characteristic parameters from the noise reduction gearbox vibration signals by the particle filter. In this paper, useing these parameters to diagnose faults by PF neural network. The diagnosis results show that the PF neural network is good. It is also confirmed that the effect of noise reduction by particle filter is good.