Ultra-short term prediction of wind power based on PSO-SVR algorithm

Aiming at the characteristics of high randomness and strong fluctuation of wind power, a support vector regression model PSO-SVR was optimized by particle swarm optimization algorithm. Taking penalty parameter c and kernel function parameter g as optimization objects, PSO was used to optimize SVR parameters, and PSO-SVR wind power prediction model was established. Based on the research and analysis of the measured data of a wind farm in northern China in 2015, the ultra-short term wind power prediction with an interval of 15 minutes is carried out based on the optimization model, and the root-mean-square error is introduced as an evaluation index. The results show that the prediction error of the optimized method is better than that of the direct SVR method. The results are compared with BP neural network and ELM algorithm. The results show that PSO-SVR model has better prediction accuracy and stability, and can improve the accuracy of wind power prediction, which has certain research value.