LS-SVM在电力系统负荷预测中的应用

Application of Least Squares Support Vector Machine to Load Forecasting in Power System

针对支持向量机(SVM)法预测电力负荷存在空间划分参数率定人为因素影响的缺陷,采用谱分析法进行周期分析,比较了二次函数趋势的年负荷序列与周期性月负荷序列,并采用最小二乘支持向量机(LS-SVMlab1.5)法预测负荷。实例结果表明,周期性的月负荷序列实测值与预测值拟合度较好,预测精度高、简捷、合理、实用。

Aiming at the shortcomings of subjective classification parameters of support vector machine for load forecasting,the cycle of load series is analyzed by using spectrum analysis method.And then it compares the trend of quadratic function for yearly load sequence with that of cyclic monthly load sequence.Finally,the LS-SVMlab1.5 software is applied to predict load.The results show that the degree of fitting between the measured and predicted value is good for cyclic monthly loading sequence,and the method is simple,reasonable and practical with higher prediction accuracy.