计及灰色关联度分析的中长期负荷灰色预测方法

Medium-and Long-term Load Forecasting Method Considering Grey Correlation Degree Analysis

中长期负荷预测是配电网规划的必要前提,预测精度的提升对于增强电力系统经济效益具有重要意义。针对传统灰色预测模型预测及拟合精度较低的缺陷,考虑到灰色关联度分析能够量化度量系统发展变化态势的特点,提出改进的灰色模型。以原始值与预测值序列之间灰色关联度最大为目标,引入炸点管理策略和自适应局部搜索策略,利用改进烟花算法实现对灰色模型背景值的权重系数和初始值修正项的优化求解。基于算例仿真结果表明,相比于传统预测模型,所提出的改进模型对于提升预测精度和拟合精度有明显效果。

Medium-and long-term load forecasting is a necessary precondition for the planning of distribution network.At the same time,the improvement of prediction accuracy is of significance to enhancing the economic benefit of powersystem.Considering that the prediction and fitting accuracies of the traditional grey forecasting model are relativelylower,an improved grey model is put forward in this paper based on grey correlation degree analysis that can quantifythe characteristics of system development and changing trend. First,the maximization of grey correlation degree be-tween sequences of the original values and predicted values is set as an objective. Then,the fried management strategyand adaptive local search strategy are introduced to improve the fireworks algorithm（FWA）. Finally,the optimizationproblem in terms of weight coefficients of the grey model＇s background value and the modified term of the initial valuesis solved. The simulation result of a numerical example shows that compared with the traditional forecasting model,theimproved model proposed in this paper has advantages in prediction and fitting accuracies.