摘要
提出了一种联合使用硬C均值(hard C-mean,HCM) 聚类算法和支持向量机(support vector machine,SVM)的电力系统短期负荷预测方法。与目前采用单一SVM的负荷预测方法相比,考虑了电力负荷变化的周期性特征,依据输入样本的相似度选取训练样本,即通过对学习样本的聚类选用同类特征数据作为预测输入,保证了数据特征的一致性,强化了历史数据规律。实际应用证明了该方法的有效性,该方法不仅提高了负荷预测精度,还缩短了预测执行时间。
A new short-term load forecasting method is proposed by conjunctive use of Hard C-mean clustering (HCM) algorithm and support vector machine (SVM). Comparing with the load forecasting method in which only the SVM is used, in the proposed method the periodical feature of power loads variation is considered, according to the similarity degree of input samples the training samples are chosen, i.e., by means of the clustering of study samples the data possessed of homogeneous characteristic are chosen and used as the input of forecasting, thus the consistency of data characteristic can be ensured and the regularity of historical data is intensified. Using this method, not only the accuracy of load forecasting is improved, but also the forecasting process is speeded up. Practical application results show that the proposed method is effective.
出处
《电网技术》
EI
CSCD
北大核心
2006年第8期81-85,共5页
Power System Technology
关键词
电力系统
短期负荷预测
支持向量机
硬C均值聚类
相似度
power system: short-term load forecasting: support vector machine (SVM): hard C-mean clustering (HCM): similarity degree