Research On Fault Diagnosis Of Wind Turbine Gearbox Based On The Integration Of Artificial Immune System And Evidence Theory

With the deeper research of the global new energy technology in recent years,the speed of development of new energy technology is getting faster,the proportion of wind industry is higher than others and the investment in wind power industry is getting more and more,hence the importance of safe operation of wind turbine is significant.It is a hot topic to study the effective methods of fault diagnosis and state monitoring of wind turbine.The wind turbine gearbox fault diagnosis method is the focus of the research.In order to make the operation of gearbox more safe and reliable,gears,bearings and some other parts are studied.Centering on the combination of negative selection algorithm optimized by genetic algorithm and improved evidence theory,forming an integrated approach to diagnose wind turbine gearbox fault.Main aspects of work are done as follow.Several common gearbox fault types and the development and current situation of fault diagnosis are reviewed.Then the principle of fault diagnosis and the framework of fault diagnosis system are introduced.And several methods for handling vibration signals and extracting feature parameters are summarized.Finally,the characteristic parameters are normalized.This paper introduces the principle and feature of artificial immune system and compares negative selection algorithm,real value negative selection algorithm and V-detector negative selection algorithm to know which one has better performance in fault detection.Then by using genetic algorithm optimized by elite strategy to optimize the process of generating detectors that generated by V-detector negative selection algorithm.For solving the problem that negative selection algorithm only can recognize whether the mechanical operation state is abnormal rather than accurately identify the type of mechanical faults,by using evidence theory to dispose the fault signal characteristic parameters to achieve the purpose of accurately diagnosing.Through experimental verification,the integration method of V-detector negative selection algorithm optimized by elite strategy combined with evidence theory can be used in gearbox fault diagnosis effectively.