Fault Diagnosis Of Wind Turbine Drive System

The wind power generation has the advantages of cleanliness,economy and safety,the installed capacity of new wind power in China is increasing year by year.In the wind power system fault,the transmission system fault has a high fault ratio,what's more,the maintenance is difficult,and the economic loss is serious.Therefore,it is necessary to detect and predict the fault of the wind power transmission sys tem.In this thesis,studying the fault diagnosis and fault prediction technology based on bearings,gears and other key components of the wind turbine drive system.Firstly,particle filter algorithm is an effective method to solve the estimation problem of non-Gaussian non-linear stochastic systems,for the disadvantage of particle dilution of particle filter,this thesis proposes an improved bat algorithm to optimize particle filter.Then,using ARMA model to determine the state equation,using FPE crite rion to determine the order of state equation,using the least square method to estimated the parameters of the state equation.Finally,the noise of the fault vibration signal is reduced and then analyzs the envelope spectrum.The experimental results sho w that the proposed algorithm greatly suppresses the interference signal in the vibration signal,and can clearly distinguish the fault frequency and classify the fault.Secondly,study traditional BP neural network's the fault identification,aiming at th e shortcomings of slow convergence speed and easy to fall into local minimum value of traditional BP neural network,this thesis optimizes the BP neural network by the improved bat algorithm,takes the root mean square error of predicted output and expecte d output as fitness function,and finds out the optimal weight and threshold of BP neural network through the improved bat algorithm,then performs fault prediction.The experimental results show that the algorithm overcomes the disadvantages of traditiona l BP neural network in fault identification,such as slow convergence rate,easy falling into local minimum value,so the fault prediction of wind turbine transmission system has good effects.