摘要
电力负荷受气象因素影响越来越大 ,如何准确预测负荷中的气象负荷是负荷预测中的一项有意义的课题。本文首先采用粗糙集对影响负荷的气象因素进行规则简约 ,找到影响负荷的核心气象因素 ;然后以这些核心因素为坐标寻找与预测日距离最小的历史数据 ,利用时间序列方法进行预测。经实际系统检验 ,证明该方法克服了传统气象负荷预测中的主观性 ,将历史数据的发掘过程量化 ,便于机器预测。并且预测结果误差小 。
Weather condition has an increasing influence upon the power load.How to forecast the weather load is an important research subject.In this paper,first Rough Set was employed to process attribute reduction on weather factors that will influence power load;this procedure can find the essential factors in power load.Then these essential factors were used as coordinates to find the nearest historical data as input for time series model.The method can avoid the subjectivity of traditional forecasting methods.It scales the match procedure of historical data so that it's easy to be applied to software realization.The forecasting error is decreased.The method is applicable to power load forecasting.
出处
《电力系统及其自动化学报》
CSCD
2004年第4期59-63,共5页
Proceedings of the CSU-EPSA
