基于负荷特性分析的短期负荷预测相似日选择方法

The Select Similar Days Method for Short-term Load Forecasting Based on Load Characteristic Analysis

从选择历史负荷数据的角度来看,各类短期负荷预测方法本质上都基于相似原理。科学合理地选择相似日是提高预测效果的有效途径。为了挖掘出负荷曲线形状与预测日最大可能相似的历史日,提出了在负荷特性分析的基础上确定日特征向量、采用模糊分类和改进灰色关联分析法进行选择的方法。预测时引入时间跨度系数以消除负荷水平差异的干扰。实际系统的应用效果验证了该方法的有效性和实用性。

As most short-term load forecasting methods are essentially relied on similarity theory respect to sample data selecting.Similar days selecting is the primary and effective way.Based on dynamic analysis on load character this paper proposed a new method for similar days selecting. The method constructs day character vector, adopts fuzzy categorization and updates grag correlation analysis. A parameter called time span compensation weight was used to improve forecasting result. The practical application by software developed verified the efficiency of this method.