日最大负荷特性分析及预测方法  

Characteristic Analysis and Forecasting Method for Daily Peak Load

作　　者：马立新[1] 李渊[1] 

机构地区：[1]上海理工大学光电信息与计算机工程学院,上海200093

出　　处：《电力系统及其自动化学报》2014年第10期31-34,共4页Proceedings of the CSU-EPSA

基　　金：国家科技部政府间科技合作项目(2009014);上海市创新基金项目(jwcxsl1302)

摘　　要：电力调度和市场营销部门对电力负荷数据的走势形态和预测方法十分重视。在实际应用中,电力市场对提前预测未来连续多天的日最大负荷提出了新的要求。本文根据电力系统中日最大负荷的历史数据,分法定假日与非假日两部分单独研究其特性。对于假日最大负荷的预测,设定假日因子;对于非假日,通过小波分解提取日最大负荷变化的周期特征,再分别建立相应的BP神经网络模型进行预测。通过对某市电力负荷数据的预测及结果表明:采用这种组合方法可行有效、预测精度满足行业要求。有较强的理论意义和广泛地应用前景。Dispatch and marketing departments of electric power put great emphasis on trend and forecasting method of power load data.In practical applications,the electricity market raises new requirement for forecasting the daily peak load in continuous multi-days ahead of time.In this paper,according to the history data of daily peak load,the characteristics of holiday and non-holiday daily peak load are investigated respectively.For the forecasting of holiday daily peak load,a holiday factor is added; for the non-holidays,the period characteristic of daily peak load is obtained through the wavelet decomposition.And then,the forecasting is respectively carried out by establishing BP neural network model.Simulation results show that this method is feasible and effective,which can meet the industry requirement for prediction accuracy and has a strong theoretical significance and wide application prospects.

关 键 词：特征提取 连续多天负荷预测 日最大负荷 假日负荷预测 神经网络 

分 类 号：TP273.22[自动化与计算机技术—检测技术与自动化装置]