| 引用本文: | 陈超,黄国勇,范玉刚,吴建德,王晓东.基于离散Fréchet距离和LS-SVM的短期负荷预测[J].电力系统保护与控制,2014,42(5):142-147.[点击复制] |
| CHEN Chao,HUANG Guo-yong,FAN Yu-gang,WU Jian-de,WANG Xiao-dong.Short-term load forecasting based on discrete Fréchet distance and LS-SVM[J].Power System Protection and Control,2014,42(5):142-147[点击复制] |
|
| 摘要: |
| 针对现有电力系统短期负荷预测精度低、数据处理量大、易受输入变量的影响等问题，提出了一种将离散Fréchet距离与LS-SVM相结合的短期负荷预测方法。分析总结了East-Slovakia Power Distribution Company提供的历年负荷数据，结合该地区的用电规律，通过引入离散Fréchet距离，建立离散曲线相似性的数学模型，选取出与基准曲线形状相似的相似日，利用相似日负荷数据对LS-SVM预测模型进行训练。经过仿真验证，并与标准LS-SVM模型得到的结果对比，所提预测方法明显提高了预测精 |
| 关键词: 离散Fréchet距离 LS-SVM 用电规律 形状相似日 短期负荷预测 |
| DOI：10.7667/j.issn.1674-3415.2014.05.023 |
| 投稿时间：2013-06-04 |
| 基金项目:国家自然科学基金资助项目（51169007）；云南省科技计划项目（2011DA005 & 2011FZ036 & 2012CA022）；云南省中青年学术和技术带头人后备人才培养计划项目(2011CI017) |
|
| Short-term load forecasting based on discrete Fréchet distance and LS-SVM |
| CHEN Chao,HUANG Guo-yong,FAN Yu-gang,WU Jian-de,WANG Xiao-dong |
| (School of Information Engineering and Automation, Kunming University of Technology, Kunming 650500, China;School of Information Engineering and Automation, Kunming University of Technology, Kunming 650500, China;
Yunnan Mineral Pipeline Transmission Engineering Technology Research Center, Kunming 650500, China) |
| Abstract: |
| In order to improve the precision of short-term load forecasting and overcome the disadvantage of large amount of data and the influence of input variables, a new method based on the combination of discrete Fréchet distance and LS-SVM is presented. This paper analyzes and summarizes the historical load data provided by the East-Slovakia Power Distribution Company. Combining with the law of the region's electricity, by introducing discrete Fréchet distance, the shape-similar days which are similar to the reference day are selected by establishing the mathematical model of discrete curve similarity, and then the similar daily load data are used to train the LS-SVM forecasting model. Through simulation and comparison with the results obtained by the standard LS-SVM model, it is proved that the prediction methods significantly improve the prediction accuracy. |
| Key words: discrete Fréchet distance LS-SVM electricity regularity shape similar days short-term load forecasting |
