Research On The State Detection And Diagnose Fault Of Wind Power Dynamotor Gearbox Based On The Vibration Signal

With the steady development of the national economy,the demand for energy continues to increase,and wind power construction has also ushered in a spring breeze.However,the opportunity always comes with challenges.The most outstanding problem is that the wind turbines frequently fail,which seriously affects the normal operation of the wind turbines,and brings some troubles to the wind power enterprise and social production work.For this,strengthening the investigation and testing of wind turbine problems can greatly reduce the related costs of wind power equipment,and is of great significance for improving the economic benefits of wind power enterprises.Main components of wind turbines mainly include blade,gearboxe,generator,yaw system,frequency converter,hydraulic system,etc.Among them,the gearbox has the highest failure rate,and its health status is directly relateds to the working performance of the whole machine.In response to this,the research of this subject focuses on the wind turbine gearbox,trying to explore how to make‘health check' and fault diagnosis the means of health status detection and fault diagnosis.The key points of this:An analysis of the geometry of the gearbox is made,which explains the necessity of fault exploration,the mechanism of the gearbox is analyzed,and the forms and causes of common faults in the gearbox are studied from the aspects of meshing frequency and signal modulation.A common diagnosis method for gearbox faults is also given in the first chapter.This paper introduces the methods that are common for vibration signal analysis.There are two main types.The first one is the time domain analysis algorithm.The method can be divided into three categories,which are time domain statistical analysis,classical modal analysis and time.The domain synchronization averaging method;the second type is the frequency domain analysis algorithm,which can be subdivided into power spectrum and envelope analysis.In addition,the algorithms such as Fourier transform and Wigner-Ville distribution are simply analyzed and explained,and their respective The advantages and disadvantages were also compared and analyzed.The wind turbine gearbox is subjected to various stresses during operation.The vibration signal generated by it is a complex non-stationary signal.In order to analyze the fault characteristics of the wind turbine gearbox more intuitively,according to the relevant characteristics of the fault vibration signal,The spectrum algorithm analyzes the acquired signal,and combines the wavelet and the cepstrum to analyze thevibration signal of the gearbox.This method enables the fault diagnosis of the wind turbine gearbox through simple and easy spectrum analysis.According to the vibration signal characteristics of the wind turbine gearbox,based on wavelet packet and BP neural network,the advantages of the two algorithms are combined to realize the fault diagnosis of the wind turbine gearbox.Among them,the wavelet packet algorithm is used to extract the eigenvector corresponding to the vibration signal in the gearbox,and then the obtained eigenvector is transmitted to the BP neural network algorithm,and the classification and training operations are implemented.Finally,the fault is judged by the analysis results,and the example is proved that the method has good feature extraction ability and fault diagnosis effect.