基于滤波算法的节假日短期负荷预测研究  

Research on Short-term Holiday Load Forecasting Based on Filter Algorithms

作　　者：陈衡[1] 王科[1] 陈丽华[1] 

机构地区：[1]西南交通大学电气工程学院,成都610031

出　　处：《电气技术》2014年第9期12-15,31,共5页Electrical Engineering

基　　金：中央高校基本科研业务费专项资金(SWJTU2011CX004EM)

摘　　要：针对节假日短期负荷变化幅度大,难以精确预测,提出了一种基于滤波算法的节假日短期负荷预测算法。简要介绍了卡尔曼滤波和维纳滤波的预测原理,结合电力负荷实际情况,建立了相应的短期负荷预测模型。对节假日负荷进行了预测,证实了应用滤波算法进行负荷预测的可行性和有效性,并对比了两种预测算法。针对大幅度负荷变化导致预测精度不高的问题,讨论了基于滤波算法误差原因,在此基础上,提出了通过引入修正因子对预测结果进行修正。修正后的预测结果具有很高的精度,证实了改进算法的正确性和有效性。When the short-term load changes greatly, it is difficult to accurately predict it. A short-term holiday load forecasting algorithm based on filter algorithm is presented. The Kalman filter and Wiener filter prediction theory are introduced briefly. Combing with the actual situation of power load, the appropriate short-term load forecasting models are established. Load is predicted for holidays,confirmed the feasibility and effectiveness of the load forecasting algorithms. And the two predication algorithms are compared. For the accuracy is not high when the loads have a great oscillation, it discusses the causes of error in the filter algorithm. On these basis, the forecasting results are modified by introducing the holiday factor. The comprehensive forecasting results have high precision, which confirm the validity and effectiveness of the advanced algorithm.

关 键 词：短期负荷预测 维纳滤波 卡尔曼滤波 修正因子 

分 类 号：TM715[电气工程—电力系统及自动化]