改进GM(1,1)模型在电力负荷预测中的应用  

Application of Improved GM (1,1) Model in Load Forecasting

作　　者：牛东晓[1] 贾建荣[1] 

机构地区：[1]华北电力大学工商管理学院,河北保定071003

出　　处：《电力科学与工程》2008年第4期28-30,共3页Electric Power Science and Engineering

基　　金：高等学校博士点专项基金(20040079008);河北省自然科学基金资助项目(G2005000584)

摘　　要：灰色预测技术原理简单、样本量小、计算简便,适合用于负荷预测,但其存在很多局限性。从建模原理出发,利用最小二乘法和积分法,改进影响模型精度最重要的两个因素——初始值和背景值。以北京市1999—2005年的电力消费量数据加以计算,证明改进效果较为显著。Gray forecasting model is suitable for load forecasting, because its principle is simple and calculation is convenient in small sample. But there are many limitations. This paper used Least Square Method and Integration Method to improve two of the most important factors of model accuracy, the initial value and the background value, based on modeling principles. The improved model showed better effect applied to the historical data of power consumption from 1999 to 2005 in Beijing.

关 键 词：灰色预测 负荷预测 最小二乘法 积分法 

分 类 号：TM715[电气工程—电力系统及自动化]