基于物元模型的电力系统中长期负荷预测

在综合考虑电力负荷预测及物元理论特点的基础上,将两者结合,提出了一种基于物元理论(可拓工程学)的中长期负荷预测方法,对物元理论在电力系统中的实际运用进行了探讨.方法首先运用逐步回归与层次分析技术确定各种因素对电力负荷的影响权重,然后利用物元理论对选中的各影响因素与电力负荷及其增长率建立物元模型,再根据系统聚类分析的方法对电力负荷及其相关环境因素的历史样本进行归纳分类,最后采用合适的物元关联函数及灰色关联函数结合未来环境因素状态对未来负荷变化模式进行识别,从而预测出电力负荷的未来值.并以我国某地区的中期负荷预测为例,说明了其有效性.

Study on the Model of Mid-long Term Load Forecasting for Power System Based on Matter Element

This paper presents a new method used in mid-long term load forecast combining each unique characteristic of power system and ME(matler element), in addition, a new thought for practical usage of ME in power system is explored. Firstly tec hniques of Stepwise Regression and Analytic Hierarchy Process are used to determ ine the weight of each factor affecting load demand. Secondly ME model are found ed with selected affecting factors,power demand and its grossing ratio by funda mental of ME. Then the features of selected affecting factors and ratio of load grossing in each classification are identified by analyzing history data with Sy stematic Cluster method. Owing to appropriate dependent functions of ME and Grey Model, the grossing pattern can be revealed by model recognition technique, thu s the load demand can be forecasted. The successful result of the application of this method in a certain area of China, proves the validity of the approach.