Study On Monitoring And Fault Diagnosis Technology For Gearbox Of Wind Turbine

With the development of science and technology, the demand of new energy isincreasing strongly for human. As the representative and primary energy of the newemerging energy technology, wind power has got more and more attention. To ensure thenormal operation of wind turbines has become a core issue of the field of wind power.Gearbox as the key equipment for the wind turbine, makes a great effort to theperformance of the wind turbine. Due to the work environment and the mechanical features,the gearbox is easy to get out of order. It is meaningful to study on monitoring and faultdiagnosis technology for gearbox of wind turbine.First, this paper introduces the composition, the working principle，and the typical faultof the gearbox, studied the mechanism of vibration signals and vibration characteristics.Signal analysis technology is studied, including Time-frequency analysis wavelet analysisand wavelet packet analysis. Wavelet packet energy feature was applied to extract theeigenvector. Second, the basic principle of particle swarm optimization algorithm andneural network wear studied. And the feasibility of their combination is studied. Theweights and thresholds of the BP neural network is optimized using particle swarmoptimization algorithm. Finally, the BP neural network of particle swarm optimization isapplied to gearbox fault diagnosis. The results of fault diagnoses have been proved to beeffective. Based on those, the design project of gearbox monitoring system is proposed.Studied in this paper on monitoring and fault diagnosis technology for gearbox is practicaland significant.