基于序列最小化算法的周负荷预测研究  

The Research on Week Load Forecasting Based on Sequential Minimal Optimization

作　　者：黄元生[1] 李孝宇[1] 

机构地区：[1]华北电力大学经济管理系,河北保定071003

出　　处：《山东电力高等专科学校学报》2013年第4期22-23,32,共3页Journal of Shandong Electric Power College

摘　　要：周负荷预测具有周规律性,对编制发供电计划有重要意义。运用序列最小化算法(SMO)求解支持向量机,既能发挥支持向量机的优势,又能简化问题,缩短运行时间。通过山西省某变电站的实例分析,得到序列最小化算法适用于求解周负荷预测,且精度较高。Week load forecasting has regularity, and is impor-tant for power system operation. The sequential minimal optimization （SMO） algorithm is used to compute load forecasting because SMO can avoid iterative, so as to short the running time. By a pow- er plant data of Shanxi Province, the result proves that sequential minimal optimization （SMO） is suit- ed to compute Week load forecasting, and it has high precision.

关 键 词：周负荷预测 支持向量回归 序列最小化算法 

分 类 号：TM715.2[电气工程—电力系统及自动化]