基于聚类分析的配电网小区负荷预测

通过引入聚类分析,对样本数据聚类,利用聚类后的结果预测小区负荷密度,并且提出了2种修正小区负荷密度的方法,有效减小了预测误差。与传统负荷密度法相比较,该方法不仅提高了预测精度,精简了优化配电网建设的资金,而且计算简单,通过算例表明其工程实用效果较好。

Small Area Load Forecasting of Distribution Network Based on Clustering Analysis

By means of drawing clustering analysis,the sampling data of all kind of areas is classified dynamically,the results of small area load density after clustering are forecasted,and two improvement spatial load forecasting methods are presented and the forecasting error is decreased effectively.Comparing with other load forecasting method,the new way not only improve the precision of the load density and optimize the investment,but also simplify the complexity of computation.The result of load forecasting computation of a certain area shows this method is effective.