改进的GM(1,1)模型在配电网日负荷预测中的应用  

Application of Improved GM(1,1) in Distribution Network Daily Load Forecasting

作　　者：陈辰旭[1] 杨耿杰[1] 郭谋发[1] 

机构地区：[1]福州大学电气工程与自动化学院,福州350108

出　　处：《电气技术》2012年第5期14-17,21,共5页Electrical Engineering

摘　　要：灰色预测法常用GM(1,1)模型,当负荷波动较大时,该模型预测精度较低。针对周期性波动较大的日负荷提出了一种基于等维新息的改进灰色预测模型,即"滑动平均—反双曲余弦"模型。该模型对等维更新的数据做预处理,充分利用预测新息降低灰度,提高离散数据的光滑度,从而提高预测精度。通过两个典型的实例介绍了改进灰色预测模型在日负荷预测中的应用。结果表明该模型能够改善预测效果。The model of GM（1.1） is usual used in grey forecasting, sometimes the prediction accuracy of the model is relatively low when the load has larger fluctuation. The paper puts forward an improved grey system model based on equal dimension and new information for the daily load which has larger cyclical fluctuation, that is ＂moving average-inverse hyperbolic cosine＂ model. In order to increase the prediction accuracy, the method do pretreatment on the data of equal dimension, it can reduce grey degree and improve the smoothness of discrete data. The application of improved grey system model in daily load forecasting is introduced through two typical and practical examples. The results show that the method can increase the forecasting accuracy.

关 键 词：日负荷预测 GM(1 1)模型 等维新息 滑动平均 反双曲余弦 

分 类 号：TM715[电气工程—电力系统及自动化]