Early warning of abnormal state of wind turbine based on principal component analysis and RBF neural network

摘要

This paper proposes an early warning method for abnormal state of wind turbines based on PCA-RBF neural network model. According to the operation and control principle of the wind turbine, determine the power as the output parameter of the model, and use correlation analysis to filter out the input parameters with strong correlation with the output power of the unit, and use the RBF and PCA-RBF methods to establish normal operating conditions Comparative analysis of the unit model. The experimental results show that the input parameters determined by the correlation coefficient are more reasonable, and the model established by the PCA-RBF neural network is more accurate than the traditional RBF neural network model. Therefore, the neural network model based on PCA-RBF is selected, combined with the sliding window method to monitor the operating status of the unit, and realize the early warning of the abnormal status of the wind turbine. The example shows that the method can give early warning in time before the failure, which proves the feasibility of the method.