Fault Diagnosis Of Wind Power Bearing Based On Order Analysis

The operating environment of the wind turbine is poor,causing frequent failure of the various components in the unit and a longer downtime due to the fault.Bearing failure caused by downtime is accounted for 60% in the all fault events.The monitor and diagnosis of bearings is of great significance to reduce the operating costs of wind turbines and improve the efficiency of wind farms.The paper mainly carried on the following several aspects of the work:In this paper,the failure mode and fault diagnosis method of rolling bearing are introduced firstly,and the fault mechanism is studied to find out the fault information and calculate the fault frequency of the bearing.The fault model of the rolling bearing is established,which provides a theoretical basis for the simulation signal.By studying the influence of the time domain and the frequency range of the vibration signal on the bearing vibration,the relationship between the signal characteristic information and the fault is found out.Secondly,the vibration analysis of the vibration signal is analyzed by the order analysis.The results show that the method has good noise reduction effect.Based on the instantaneous frequency of the STFT transform peak signal,the rotational speed of the rotating shaft is estimated and compared with the rotational speed signal.The simulation results show that the method has high accuracy in estimating the rotational speed of the rotating shaft.Finally,the power spectrum of the vibration signal is extracted as fault feature information for neural network fault diagnosis.Finally,the structure and algorithm of BP neural network are studied,and the fault diagnosis model is established according to the fault feature information.A large number of sample data are needed for neural network learning and training,and a WTDS experimental platform for rolling bearing failure is established.The vibration data of bearing is acquired to ensure the accuracy and accuracy of fault diagnosis.Based on the shortcomings of traditional BP algorithm,an adaptive dynamic learning rate algorithm is designed.Compared with the traditional improved algorithm,the simulation results show that the new algorithm is effective and accurate in the bearing fault diagnosis.The rolling bearing fault diagnosis system is developed by the LabVIEW software,and the demand of online monitoring and fault diagnosis is realized.The simulation results show that the system can accurately diagnose the bearing failure.