IEGABP fault prediction model for wind power gearbox

Gearbox is the key component of large scale wind turbines, should adopt suitable forecasting model to predict the state. Genetic algorithm optimization BP neural network prediction mode based on information entropy is proposed. This model can be used to optimize the parameters of neural network weights, and improve the prediction accuracy and timeliness with the use of the contribution of different vibration data to prediction. Vibration data of wind turbine gearbox is acquired on site, and IEGABP model and BP neural network structure parameter model of artificial experience are used to predict and compared the result, the results show that the former has achieved good results in the prediction accuracy, prediction of real-time.