基于偏最小二乘回归分析的短期负荷预测

SHORT-TERM LOAD FORECASTING BASED ON PARTIAL LEAST-SQUARES REGRESSION

对偏最小二乘回归分析在电力系统短期负荷预测中的应用进行了研究。该方法可有效地进行数据准备和样本预处理,并可以对输入因素进行成分提取。提取出的成分具有线性无关的特点,对日负荷有较好的解释能力,且利于建模和预测。此方法另一特点是可以消除输入因素的多重共线性,不需要大量样本作为输入。算例表明,该方法用于短期负荷预测建模速度快、预测精度高,是一种行之有效的方法。

The effectiveness of applying partial least squares regression to short-term load forecasting in power system is investigated. This algorithm not only can effectively perform the data preparing and samples pretreating, but also can abstract components from input factors. The abstracted components, having the property of linearly independence, can be used to explain the daily load better and are more favorable for modeling and forecasting. Another feature of this method is that the multicollinearity of input factors can be eliminated, so a lot of samples to input are not needed. The results of calculation example from a real power system show that in short-term load forecasting the presented method is effectual, with this method the high modeling speed and high forecasting accuracy can be obtained.