Research On Fault Diagnosis Methods Of Planetary Gear Box For Wind Turbine

Recently,wind turbines often fail with the completion of wind power in a large-scale.Therefore,the research on the fault diagnosis of the planetary gear box of the wind turbine generator is of great significance to the maintenance of the normal operation of the unit and the reduction of the economic losses.At present,the fault diagnosis analysis of the planetary gearbox is mainly focusing on the analysis of the vibration mechanism of the fault and the frequency spectrum of the vibration signals.In this paper,a method of planetary gear fault pattern recognition which is based on the characteristic waveform is proposed and then a gearbox fault diagnosis method based on multi feature pattern recognition is developed to recognize the fault patterns of planetary gear box.The main contents are as follows:1.Research on the fault type and vibration mechanism of wind turbine gearbox.In this paper,the reasons why the wind turbine gearbox is easily to have faults are introduced from the aspects of the composition of the wind turbine gearbox and the working principles.And then,the gearbox fault types of wind turbine and the corresponding causes of prosperity are analyzed deeply.At the same time,the mechanism of the fault and the frequency characteristic of the vibration signal of the gear fault are analyzed on the part of the fault mechanism of gearbox.2.Planetary gear fault pattern recognition method based on characteristic waveform.This planetary gear fault pattern recognition method of feature waveform is extracted the feature waveform based on the planetary gear fault vibration signals.Then,a step is used to optimize the fault feature waveform and the optimized feature waveforms are used for planetary gear fault pattern recognition.For the reason that the proposed method has no characteristic formula,the fault feature extraction is simplified greatly and the computing time is reduced in a large scale.Thus,it improves the identification efficiency and can realizes gear fault pattern recognition accuracy.Last,the effectiveness of the proposed fault pattern recognition method is verified with the fault pattern recognition of the experimental signals in the planetary gearbox.3.Gear box fault diagnosis based on multi feature pattern recognition.A multi feature extraction framework is built aiming at the problem that the traditional methods are difficult to extract the fault feature of gear effectively.In this method,it extracted RMS,peak index,kurtosis index,pulse index in the time domain and make variational mode energy entropy as feature vectors of fault identification.Then,the recognition method of support vector machine is used to classify the fault.In order to solve the problem that the SVM model parameters have a great influence on the classification results,a particle swarm optimization(PSO)algorithm is proposed to optimize the C and G parameters.On this basis,the experimental data are used to verify the method.The results show that the proposed recognition algorithm has higher recognition rate,with which the validity of the proposed method is verified.