Research On Intelligent Fault Diagnosis Method Of Wind Turbine Gearbox

As a kind of high yield,renewable and no pollution energy,wind energy gained wide attention from countries all over the world.However,with the continuous increase of wind turbine capacities,the failures and accidents of wind turbine are also increased,which seriously affects the normal electric power generation.Therefore,in order to ensure the normal and stable operation of wind turbine,it is necessary to do the diagnosis work.As an important transmission part of wind turbine,the failure rate of gearbox is very high.Once failure occurs,the maintenance work will be very difficult.Therefore,this paper does some deep researches of the fault diagnosis work of wind turbine gearbox.The main work of this paper are as follows:(1)Study the common faults of wind turbine gearbox.Understand the causes of the faults and analyze the vibration signals of gearbox under different running conditions.(2)An improved threshold wavelet denoising method is proposed,which is used to reduce the noise of vibration signals;extract the characteristic value of signals under different operation states from time domain,frequency domain and node energy.Then,construct the sample data set.(3)Proposed the ISALAFA algorithm and use it in the parameters selection of extreme learning machine.After building the ISFLAFA-ELM model,use it in the fault diagnosis work of bearing;proposed ISAPSO algorithm and use it in the parameters selection of support vector machine.After building the ISAPSO-SVM model,use it in the fault diagnosis work of gear.The effectiveness in engineering application of these two methods areverified by the use of actual wind turbine operation data.(4)Design a software system for fault diagnosis of wind turbine gearbox.With the combination of Labview and Matlab,the system can realize signal processing,characteristic value extraction and the fault diagnosis work of wind turbine gearbox.It makes diagnosis work more intuitive and simple.