基于偏最小二乘法的支持向量机短期负荷预测

摘要

This paper presented a short -term load forecasting model based on partial least square and SVM, firstly drawing loading data compenont by partial least square (PLS), the compenont have the linearity irrelative characteristic, and eliminating multi -relative of input factor subsequently. And then using support vector machines (SVM) achieve load forecasting. The example indicate that the presented short-term load forecasting model comply with fast model and forecast accuracy, which is an effective method%提出了一种基于偏最小二乘支持向量机的负荷预测模型.首先通过偏最小二乘(PLS)对负荷数据进行成分提取,提取的成分具有线性特点,并消除输入因素的多重相关性,然后采用支持向量机方法(SVM)对提取的成分进行预测.算例表明,该算法用于短期负荷预测建模速度快,预测精度高,是种行之有效的方法.