基于双修正因子的模糊时间序列日最大负荷预测  

Maximum Load Forecasting Based on a Bi-Factor Revised Fuzzy Time Series Model

作　　者：刘晓娟[1,2] 方建安[1] 

机构地区：[1]东华大学信息科学与技术学院,上海201620 [2]上海电力学院数理学院,上海201300

出　　处：《中国电力》2013年第10期115-118,共4页Electric Power

基　　金：国家自然科学基金青年项目(61203006)

摘　　要：天气温度变化是影响短期电力负荷预测的主要因素。为提高预测精度,引入负荷变化影响因子和气温影响因子,提出基于双修正因子的模糊时间序列预测算法。根据负荷变化趋势,提出分段预测的思想,在拐点处用负荷变化因子进行修正,然后用气温影响因子对预测结果进行二次修正。将改进的算法用于某电网夏季最大负荷的预测,数值结果表明该算法具有较高的预测精度。Weather temperature is the main factor to affect the short-term power load forecasting.In order to improve the accuracy of the forecast,a bi-factor revised fuzzy time series model is proposed for maximum load forecasting.The influence factors of power load trend and temperature are introduced into the conventional fuzzy time series forecasting algorithms to correct the forecasting results.The segmented prediction idea is proposed in accordance with the trend of the load changes.Correction is made at the inflection point by load trend factor,and temperature influence factor is used for secondary correction on the predicted results.The model was applied to the Suzhou Grid for the maximum load prediction in summer and the results show that the model has a better prediction accuracy.

关 键 词：电力系统 负荷预测 模糊时间序列 负荷变化影响因子 气温影响因子 

分 类 号：TM715[电气工程—电力系统及自动化]