基于最小二乘支持向量机的超短期风电负荷预测

风力具有很强的间歇性和波动性,导致风电负荷预测困难,主要表现在预测计算速度慢,可预测的未来时间短,预测精度不高。为了解决这些预测困难,将最小二乘支持向量机(LS-SVM)的方法运用在超短期风电负荷预测中。最小二乘支持向量机通过改进算法,简化了计算的复杂性,使计算速度明显增快,也进一步提高了预测的精度。用实际数据进行仿真,实验结果表明,基于LS-SVM的方法可以进一步提高超短期风电负荷预测的精度,加快计算和预测的速度,与其他方法相比预测精度和运算速度都有优势,用于超短期风电负荷预测是有效可行的。

Ultra-Short-Term Wind Power Load Forecast Based on Least Squares SVM

Strong intermittency and fluctuation of wind leads to difficulties in wind power load forecast,such as slow forecast and calculation, short predictable future and low prediction accuracy.To overcome these difficulties,the least squares support vector machine （LS-SVM）method is used in the ultra-short-term wind power load forecast.The improved LS-SVMcalculation simplifies the computational complexity,raises computation speed remarkably,and improves prediction accuracy.Results of simulation made with actual data show that the method based on LS-SVM can further improve prediction accuracy of ultra-short-term wind power forecast and raise calculation and prediction speed,showing advantages both in prediction accuracy and calculation speed as compared with other methods.This method is feasible and effective when used for ultra-short-term wind power load forecast.