Research On Fault Diagnosis System Of Wind Power Gearbox Based On Fuzzy Neural Network

The wind power gearbox is one of the most important equipment in the wind power generation system.Once it fails,it will cause serious losses to the whole wind power generation system.Therefore,it is of great significance to study the fault diagnosis technology of the wind power gearbox.In order to be able to monitor the operation status of the wind power gearbox in real time and diagnose the faults in time,this paper aims at the existing problems of the fault diagnosis of the wind power gearbox,studies its fault mechanism in depth,and further studies the fault diagnosis algorithm of the gearbox on this basis,and designs the fault diagnosis and monitoring system of the wind power gearbox.The research work in this paper includes:(1)Research on the failure mechanism of wind power gearbox.Firstly,the working principle,common fault types,fault generation mechanism and fault characterization of the wind power gearbox are studied in depth.Based on the fault mechanism analysis of the gearbox,different types of faults are summarized,and different types of fault feature information are collected and analyzed.Lay the foundation for the follow-up fault diagnosis work.(2)In view of the traditional fault diagnosis methods and the fault diagnosis methods based on mechanism model are difficult to meet the fault diagnosis requirements of wind power gearbox,this paper uses the fault diagnosis method based on data mining to carry out the research of gearbox fault diagnosis algorithm.Based on the study of gearbox fault mechanism,a fault diagnosis model is established by using fuzzy neural network,and considering the characteristics of vibration signal of wind power gearbox,six characteristic variables of vibration signal,power spectrum entropy,wavelet entropy,correlation dimension,box dimension,skewness and kurtosis,are extracted as the input variables of neural network diagnosis model.The simulation experiment based on Matlab shows that the fault diagnosis algorithm of wind power gearbox using fuzzy neural network has higher diagnosis accuracy than BP and RBF network.(3)In order to further improve the accuracy of the fault diagnosis algorithm of wind power gearbox,the global search performance of particle swarm optimization(PSO)algorithm is used to overcome the problems of slow convergence speed and insufficient generalization ability when using gradient algorithm to train fuzzy neural network.Meanwhile,in order to further improve the optimization ability of PSO algorithm,an IPSO-FNN fault diagnosis algorithm based on improved particle swarm optimization is proposed.PSO algorithm is improved from differential evolution,inertia weight adjustment and learning factor adjustment,which improves the diversity of particle swarm and balances the global search and local development ability of the algorithm.Simulation results show that the proposed method has higher convergence speed and diagnosis accuracy than the fuzzy neural network and PSO algorithm.(4)Based on the research of fault diagnosis algorithm of wind power gearbox,the fault diagnosis and monitoring system of wind power gearbox is designed by using VS2015 software.It mainly includes four modules: user login,data processing,simulation,result query,etc.the background running program is compiled by combining Matlab software and SQL Sever database to complete the construction of fault diagnosis and monitoring system.