智能用电用户行为分析的聚类优选策略

针对大数据背景下用户智能用电行为最佳聚类数目的选择问题,提出一种用户用电行为分析的聚类优选策略。在前期智能用电用户行为分析的特征优选策略研究的基础上,采用特征优选策略提取负荷曲线的最佳特征集对用户用电数据进行聚类分析;然后提出聚类数优选策略,通过综合考虑准确度评价指标和有效度评价指标确定最佳聚类数目。以国内外的用电数据为数据源,仿真验证了所述策略可以选择合理的聚类数目,有效提高用电行为分析的数据聚类效果。

Clustering Optimization Strategy for Electricity Consumption Behavior Analysis in Smart Grid

Aiming at the problem of choosing the optimal cluster number for electricity consumption behavior in smart grid under the background of big data, a clustering optimization strategy of electricity consumption behavior analysis is put forward. Based on the research of the feature optimization selection strategy of the electricity consumption behavior analysis in the early stage, the feature optimization selection strategy is used to extract the optimal feature set of the load curve to cluster the users'' electricity data. Then the optimal strategy of clustering number is proposed, and the optimal number of clusters is determined by comprehensively considering the accuracy evaluation index and the effectiveness evaluation index. Based on the data of domestic and foreign electricity data, the experimental simulation verifies that the proposed strategy can select the reasonable number of clustering and effectively improve the clustering effect of the data of electricity consumption.