基于PSO-GA-SVM的风电功率短期预测

Short-term Forecasting of Wind Power Based on GPSO-A-SVM

风电功率精确预测是电力系统稳定运行的必要条件之一。文章以中国某一地区收集到的风速数据为基础,通过一定的分析方法建立了SVM风电功率预测模型,采用以粒子群为优化算法的PSO-GA混合算法来优化SVM预测模型。通过分析SVM、GA-SVM、PSO-GA-SVM三种模型的预测结果,并与实验数据相比较,仿真表明PSO-GA-SVM模型预测精度更高,PAO-GA混合算法优化效果更好。

Accurate forecasting of wind power is one of the necessary conditions for stable operation of the power system.Based on wind speed data collected in a certain region of China,this paper establishes the SVM wind power forecasting model through certain analysis methods,and adopts PSO-GA hybrid algorithm with particle swarm optimization algorithm to optimize the SVM forecasting model.By analyzing the forecasting results of three models of SVM,GA-SVM and PAO-GA-SVM,and comparing with the experimental data,the simulation results show that the PAO-GA-SVM model has higher forecasting accuracy and the PAO-GA hybrid algorithm has better optimization effect.