| The precision of wind power prediction is mainly influenced by the precision of the numerical weather prediction,the preprocessing method of modeling sample data,the precision of the prediction model and so on.With the maturity of predictive modeling technology,the accuracy of wind power prediction is greatly influenced by the accuracy of numerical weather prediction and data preprocessing method.On the other hand,due to the lack of comprehensive assessment system,the evaluation indexes of prediction are decided from the error,which is the single index for prediction assessment,without considering the power generation plan and real-time scheduling information.This assessment system cannot reflect the impact of randomness and fluctuation of wind power.Therefore,it is necessary to standardize the prediction error assessment system,so as to provide basis for the prediction system assessment of dispatching department.In this paper,the functional modules of the wind power prediction system based on the exclusive numerical weather prediction are studied in detail which is shown as follows.(1)In the aspect of the precision of the numerical weather prediction,this paper studies the weather prediction,the key original data of wind power prediction.Considering the proprietary data of wind power industry,an exclusive numerical weather prediction model is set up in order to improve the prediction accuracy and resolution ratio of numerical weather prediction in the vicinity of wind farms.(2)In the aspect of preprocessing method for modeling sample data,considering the lack of wind data,a data preprocessing method based on virtual wind measurement mast is proposed to provide complete data sample for prediction modeling.(3)In the aspect of prediction modeling,considering the complete high-precision wind data,a power prediction model based on the RBF neural network is built to realize the wind power prediction and wind speed prediction of wind measurement mast.(4)In the aspect of prediction error assessment system,the general error analysis method is studied.Then the cause of the error is analyzed,which helps establishing the prediction error assessment system,providing basis for power prediction assessment.(5)In this paper,a wind power prediction system is designed to realize exclusive numerical weather prediction,wind power prediction and prediction error assessment,aiming at providing a set of software system which is convenient for operation and management.It can not only provide help for wind power prediction and assessment but also make the power scheduling plan more scientific,which can enhance the wind power consumption and improve power efficiency. |