Research On Wind Power Interval Prediction Based On Copula Function

Wind power has strong randomness and volatility,so it is very difficult to perform real-time scheduling using only point prediction technology.As a forecasting method,interval prediction can calculate the upper and lower limits of wind power at a given confidence level,which is of great significance to ensure the safe operation of power systems.At present,a large number of scholars have conducted research on the problem of wind power interval prediction and formed a rich prediction theory,but most methods need to assume that the wind power follows a certain regression model,failing to consider the correlation between wind power sequences in adjacent periods,and Or the corresponding model parameters need to be set in advance during the prediction process,which reduces the prediction accuracy of wind power to a certain extent.In view of the above problems,this paper introduces the conditional Copula function to study the problem of wind power interval prediction.Firstly,an interval prediction method based on the conditional Copula function in discrete form is proposed.This method can make full use of the excellent characteristics of the Copula function.By establishing the discrete conditional Copula function of the point to be predicted,the correlation between the wind power sequences in adjacent periods is mined.Obtain a prediction interval of the point to be predicted,and then use rolling prediction to obtain the prediction result of the entire prediction period.This method is applied to three wind farms and compared with ANN(Artificial Neural Network)and ARMA(Autoregressive moving average model),PICP(prediction interval coverage probability)is more High,PIAW(prediction interval average width)is narrower,and the prediction effect is more excellent.In order to solve the problem of selecting parameters(K,t)in the interval prediction process of the Copula function,a multi-objective interval prediction model based on the conditional Copula function is proposed.The model is within the value range of the Copula interval prediction model parameters,using PICP and PIAW For the objective function,the NSGA-II(Non-dominated Sorting Genetic Algorithm-II,second-generation non-dominated sorting genetic algorithm)multi-objective optimization algorithm is used to find the non-inferior solution set of the prediction model parameters,and then the entropy weight method is used to determine the PIAW and The respective weights of PICP are calculated by weighting to obtain a set of optimal prediction model parameters,and the Copula interval prediction is carried out with this parameter,so that the interval prediction effect is further improved.Finally,in order to fully consider the correlation between wind power and weather factors,a multivariate conditional Copula function multi-objective wind power prediction interval prediction method is proposed.This method combines historical weather data and power data to establish a conditional Copula function.The conditional Copula function of multivariate distribution thus characterizes the correlation between various factors.In the calculation example,this method shows better applicability and effectiveness than ANN and ARMA.