组合预测模型在电力系统负荷预测中的应用研究  

Application of Combined Medium and Long-term Load Forecasting Model on Power System

作　　者：江朋[1] 童述林[1] 万志宏[1] 

机构地区：[1]华南理工大学电力学院

出　　处：《自动化与信息工程》2011年第1期46-48,共3页Automation & Information Engineering

摘　　要：本文基于灰色预测模型、滑动平均模型和指数平滑模型这三种单一预测模型,采用方差-协方差策略,建立组合预测模型。然后结合老挝电力系统的概况,对老挝的全国年用电量进行预测和分析。结果表明,组合预测模型的预测精度明显高于各单一预测模型,即组合预测模型的相对误差小于各单一预测模型的相对误差,说明组合预测模型具有相当的适用性和优越性。A novel combined medium and long-term load forecasting model is presented, which based on three different medium and long-term load forecasting model - gray forecasting model, moving average model and exponential smoothing model. The Variance-Covariance Approach is used to gain each model’s weighting factor. According the power system of Laos, the annual electricity demand is forecasted by the three single and combined forecasting models, respectively. The results showed the forecasting precision of combined forecasting models is better than each of the three single forecasting models, and the relative error of combined forecasting models is lower than each of the three single forecasting models. The combined forecasting model shows considerable applicability and superiority.

关 键 词：中长期负荷预测 年用电量 组合预测 老挝电力 

分 类 号：TM715[电气工程—电力系统及自动化]