用电量数据的可视化研究综述

可视化技术是挖掘用电量数据中所蕴含丰富信息的关键技术手段，其相关研究热点集中在四个方面。用电量可视化分析形式主要有堆叠面积图等标准图表和一些自定义图表。用电量监控的可视化系统常用于建筑群和工业过程的能耗分析，其可以在用电行为纠正等方面发挥作用。用电异常的可视化包括异常计量用电数据和异常用电行为两个方面，以从源头提高计量数据质量，减少防窃电工作量。用电量数据挖掘的可视化研究集中在用电量预测中更实时的交互手段和用电用户分类中聚类分析算法的过程展现。以往各项的研究仍存在数据量较小、数据维度不高、未考虑外部因素等不足。未来应提升数据量级、扩大时空范围、融合多源数据、建立更细颗粒度用电数据模型。

Overview of visualization research on electricity consumption data

Visualization is one key technology to mine the rich information contained in electricity consumption data. Its related research hot spots are mainly display in the following four areas. The visual analysis forms mainly include standard charts such as stack area and some custom charts. The visualization monitoring system is often used to analyze the energy consumption of buildings and industrial processes, and it can play a role in the correction of electricity consumption behavior. Abnormal electricity consumption visualization includes abnormal metering data and abnormal consumption behaviors, so as to improve the quality of metering data from the source and reduce the workload in preventing electricity theft. The electricity consumption data mining visualization focuses on more real-time interactive matters in electricity consumption prediction and the presentation of a clustering analysis process in user classification. There are some shortcomings in previous studies, such as small data size, low data dimension and no consideration of external factors. In the future, the data magnitude and spatio-temporal range should be improved, multiple source data should be fused and an electricity data model with finer granularity should be established. This work is supported by the National Key Research and Development Program of China (No. 2020YFB0906004).