基于小波支持向量机的电力系统短期负荷预测

根据相空间延迟坐标重构理论,基于支持向量机强大的非线性映射能力和小波核函数的局部分析和特征提取能力,提出了一种基于小波支持向量机的电力系统短期负荷预测方法,并利用该方法对嵌入维数与预测性能的关系进行了探讨。仿真结果表明,该预测方法能精确地预测电力负荷,而且在电力负荷序列的最佳嵌入维数未知时也能取得比较好的预测效果,这一结论预示着小波支持向量机是一种预测电力系统短期负荷的有效方法。

Short-term Power Load Forecasting Based on Wavelet Support Vector Machine

Based on the powerful nonlinear mapping ability of support vector machines and the ability of wavelet kernel function in analyzing locally and extracting characteristic features from time series,the short-term power load forecasting method of wavelet support vector machines(WSVM) in combination with delay coordinate phase reconstruction of chaotic time series is proposed.The relationship between the embedding dimension and Root Mean Square Error of this model is discussed.Simulations show that the WSVM can predict short-term power load accurately,and even if the optimal embedding dimension is unknown,the predicted results are satisfactory.This result implies the WSVM is a good tool to forecast short-tern power load.