摘要
短期电力负荷预测中,针对维数比较高、各影响因素差异大、随机误差差异性大等问题,提出一种基于加权相似度和加权支持向量机的模型。首先,通过主成分分析得到负荷数据的综合因子,利用灰色关联分析分析综合因子与各影响因素的关系,计算各个影响因素的权重;其次依据权重采用加权相似度公式获得相似日,即样本数据;最后,针对相似日,采用加权支持向量回归机进行建模,实现对短期电力负荷进行预测。实验结果表明了该方法的有效性。
As for forecasting short-term power load, relatively high dimension, and large differences in random error and its fac- tors, this paper is based on weighted similarity and weighted support vector machine model for forecasting short-term power load. Principal component analysis is consolidated to get load data factor, and the use of gray relational analysis and computation of a composite factor of the relationship among various factors helps to arrive at the weight of various factors. According to the weight, the weighted similarity formula can conclude similar day, i. e. the sample data. For similar days, using weighted sup- port vector regression model ultimately can make the short-term power load forecasting. The results show that the method is an effective short-term power load forecasting.
出处
《计算机工程与设计》
CSCD
北大核心
2012年第12期4769-4773,共5页
Computer Engineering and Design
关键词
主成分
灰色关联分析
相似度
加权支持向量机
负荷预测
principal component
gray correlation analysis
similarity
weighted support vector machine
load forecasting
