Research On Fault Diagnosis Of Wind Turbines Based On Quantum Evolutionary LS-SVM And Bayesian Probability Analysis

As a new energy power generation method with mature technology,large development scale and good commercial development,wind power generation has become the fastest growing green energy in the world.However,the geographical environment in which wind turbines operate is complex,and most of them are in deserted areas.With the increasing growth of the wind power industry,more and more time and economy are spent on fault diagnosis of wind turbines,which poses a great challenge to the reliability of wind turbines,so timely and effectively fault diagnosis of wind turbines,improving the reliability of wind turbines,and reducing the cost of wind turbines operation and maintenance have become critical issues to be solved urgently.In this paper,the wind turbines fault data obtained by FAST simulation software is used to study the wind turbines fault diagnosis through the proposed fault diagnosis method based on quantum evolution least squares support vector machine and Bayesian probability analysis.(1)This article first briefly describes the basic structure and working principle of the wind turbines,and analyzes the faults of the wind turbines.The basic theory of FAST simulation software is introduced,simulates the fault of wind turbine through FAST,explains the generation of fault in detail,and obtains the fault data needed in the research.(2)Aiming at the problems of various uncertainties and noise on data,a fault recognition model based on quantum evolution least squares support vector machine is proposed.A least squares support vector machine is used to establish a fault recognition model for wind turbines.A quantum evolution algorithm is introduced to optimize the regularization parameters and kernel function parameters.The bit coding method and quantum revolving gate update strategy in quantum evolution are used to improve the diversity of the population.By comparing the obtained fault data with the traditional LS-SVM and PSO-LS-SVM methods,it is proved that the fault recognition model based on quantum evolution least squares support vector machine has good effect.(3)Wind turbines is a complex system with multiple components coupled,and there is a certain degree of coupling in the occurrence of faults.Although most of the previous methods of fault diagnosis for wind turbines can judge the fault of key components of wind turbines,The results obtained are unique and exclusive,that is,only a single,deterministic fault conclusion can be obtained,which deviates from the habit of on-site technicians for maintenance and repair operations of wind turbine generator equipment.Moreover,once a fault false alarm occurs,other possible alternative faults cannot be given.For the above problems,this paper introduces Bayesian probability formula on the basis of wind turbine fault identification to analyze the fault identification results,and gives each one the probability of fault occurrence is used as the result of fault diagnosis,and the troubleshooting sequence is given.