计及气象时空分布影响的短期负荷预测方法  

Short-Term Load Forecasting Methods with the Impact of Temporal and Spatial Distribution of Weather

作　　者：吴锋[1] 胡剑琛[1] 余加喜[1] 刘充许[2] 

机构地区：[1]海南电网公司,海口570203 [2]清华大学,北京100084

出　　处：《南方电网技术》2012年第5期79-82,共4页Southern Power System Technology

摘　　要：短期负荷受气象变化影响明显,进行预测时需要考虑气象信息的时空分布特征。分析了气象时空分布特征对负荷产生的影响,采用相关性分析方法筛选出与负荷相关性较强的气象要素,并分析了负荷受降雨影响在时间维度上的滞后特性。研究了计及气象时空分布特性的相似日法、人工神经元网络法、决策树法等预测方法。The result of short-term load forecasting is obviously influenced by weather, and thus the temporal and spatial distribution characters of weather need consider in the forecasting procedure. Analyzing the impact of the temporal and spatial distribution of weather, this paper screens out the characteristic elements of larger relativity to the short-tern load with correlation analysis methods, and analyzes the temporal hysteresis of rain influence on the load. The similar date method, artificial neural network method, and decision tree method, all taking the the temporal and spatial distribution characters of weather into account, are also studied.

关 键 词：短期负荷预测 气象时空分布 人工神经元网络 决策树 

分 类 号：TM714[电气工程—电力系统及自动化] P458.2[天文地球—大气科学及气象学]