基于遗传优化的支持向量机短期风电功率预测研究

A research of SVM short-term wind power prediction based on GA optimization

风电具有波动性、间歇性、随机性等弊端,故而较为准确的预测风电功率是提高电力系统安全性与经济性的重要手段。利用遗传算法对支持向量机参数寻优,据此建立功率预测模型进行仿真,最后与标准支持向量机的预测结果进行对比,结果表明该预测方法在短期风电功率预测中准确性更高。

According to the characteristics of volatility, intermittent and randomness of wind power, fairly accurate forecasting of wind power is an effective means for improving security and economy of the power system related. In this paper, GA is applied to optimize the parameters of SVM. Then a forecast model is developed to simulate and predict the wind power. Finally, the results show that, compared with the predicting outcomes of original SVM, the optimized SVM based on GA is found to gain a high accuracy of wind power prediction.