基于粗糙集理论与D-S证据理论改进的多元回归负荷预测方法研究

发布时间：2018-01-11 21:24

本文关键词：基于粗糙集理论与D-S证据理论改进的多元回归负荷预测方法研究　出处：《电力系统保护与控制》2016年06期 　论文类型：期刊论文

【摘要】：当前,中长期负荷预测大多采用多元回归算法,但在建模时对影响因子及历史年的选择缺乏良好的依据,很难在考虑更多影响因子及历史年数据与降低回归模型误差之间做出平衡。这使多元回归算法在实际负荷预测中的精准度很不稳定。将粗糙集理论与D-S证据理论引入多元回归算法,利用粗糙集理论对影响因子进行重要性排序。分别以历史年和影响因子为对象进行聚类,以此建立多个多元回归模型。利用D-S证据理论对多个组合预测的权重分配方案进行权重融合,得出最终基于多元回归分析法的组合预测模型。经算例验证,该模型能较好地平衡影响因子和历史年的选取,能有效提高多元回归算法在中长期负荷预测中的准确性,适用性强。
[Abstract]:At present, the medium and long term load forecasting mostly adopts multivariate regression algorithm, but in modeling, there is no good basis for the choice of influence factors and historical years. It is difficult to make a balance between considering more influence factors and historical data and reducing the error of regression model. This makes the accuracy of multivariate regression algorithm very unstable in actual load forecasting. The rough set theory and D-S evidence are combined. Multivariate regression algorithm is introduced in this paper. The importance of the influence factors is sorted by rough set theory. The historical years and the influence factors are taken as the objects for clustering. Using D-S evidence theory, the weight distribution scheme of multiple combination prediction is fused, and the combined prediction model based on multiple regression analysis is obtained. The results are verified by an example. The model can balance the factors of influence and the selection of historical years, and improve the accuracy of multivariate regression algorithm in medium and long term load forecasting.
【作者单位】：国网湖南省电力公司通信公司;湖南大学电气与信息工程学院;
【分类号】：TM715
【正文快照】： 0引言电力系统中长期负荷预测中负荷受经济、社会和气象等不确定因素影响较大,如何从多因素中提取关键影响因子和减小多因素综合所带来的不确定性并提高预测结果的精确性成为亟待解决的问题。目前常用的中长期负荷预测算法中,偏最小二乘回归预测模型对影响因子的利用比较有效,

【相似文献】