Research On Fault Diagnosis Of Misalignment For Transmission System Of Doubly-Fed Wind Turbines

In recent years,with the reduction of traditional energy sources such as coal and the increasing environmental pollution,wind energy has been developed rapidly as a clean,safe and renewable energy source.At present,wind farms mainly use the large capacity variable speed constant frequency(VSCF)doubly-fed wind turbines(DFWT)as the main models,and misalignment of the transmission system of the wind turbine is one of the main factors that causes the faults.When the misalignment occurs,it is easy to cause the damage to gear box gears,bearings and generator bearings and so on,and make the noise increase at the same time.Therefore,the mechanism analysis,fault prediction and diagnosis of the transmission system of the wind turbine is the key to ensure its long-term stable operation and safe power generation.This project is from the National Natural Science Foundation of China,named"Analysis of failure mechanisms and fault diagnostics under eccentric error for DFIG wind turbines"(Grant No.51577008).This article focus on the transmission system of 1.5MW doubly-fed wind turbine in a factory,and vibration characteristics,fault feature extraction and fault diagnosis methods of misalignment of the transmission system are studied.The main job and innovation are as follows:(1)The background and significance of the chosen topic are discussed.The research status of fault diagnosis information acquisition,fault feature extraction and fault pattern recognition of the transmission system of wind turbine are reviewed.Based on the above analysis and summary,the research contents of the dissertation are presented.(2)The gear box,coupling and generator of the transmission system are analyzed.According to the structural parameters of the components,the dynamic modeling and simulation are carried out by SolidWorks and ADAMS,and the key features of the transmission system are compared with the theoretical calculation results to verify the correctness of the model.(3)The parallel misalignment,angular misalignment and comprehensive misalignment of the transmission system are simulated by ADAMS,and the process of misalignment that changes from scratch,from small to large is simulated.The time domain and frequency domain curves of radial and axial force are extracted,and the vibration characteristics of the three kinds of faults are analyzed.Compared with theoretical analysis,the results show the feasibility of using simulation to establish faults.(4)Aiming at the traditional fault extraction methods that only a single feature or single domain feature is extracted without taking into account various features,a new mixed fault feature library construction method is proposed.The characteristic signals of misalignment of transmission system are gotten through lots of simulation,and extract the time domain and frequency domain features,and meanwhile extract the improved empirical mode decomposition(IEMD)energy entropies as the time-frequency features.Build the mixed fault feature library with the three types features,and obtain more accurate and comprehensive information that reflect the characteristic signals.(5)The support vector machine(SVM)is used as the classifier,and the particle swarm optimization(PSO)algorithm is used to optimize the parameters of SVM.The mixed fault feature library is taken as the input of SVM,and the types of misalignment are classified and identified successfully,and compared with other methods,the accuracy of fault diagnosis is improved.