Understanding multi-scale spatiotemporal energy consumption data: A visual analysis approach

Abstract

Understanding energy consumption patterns is crucial for energy demand-side management. Unlike traditional data mining or machine learning-based methods, this paper presents visual analysis methods for exploring energy consumption data from spatial, temporal, and spatiotemporal dimensions, including variability, segmentation, and energy demand shifts. To support the proposed methods, we develop a visual analysis tool that allows users to explore consumption data and validate their hypotheses based on visual results through human–client–server interactions. In particular, we propose a novel potential flow-based method to model energy demand shift patterns and have integrated it into the proposed analysis tool. We comprehensively evaluate the proposed methods and the tool using real-world electricity consumption data from the Shanghai Pudong district, and compare with traditional data mining methods. The results demonstrated the effectiveness and superiority of the proposed visual analysis methods, including its ability to discover the spatiotemporal variability of energy demand, customer groups, and demand shift patterns across different geographical areas and time horizons. All results can be well explained by knowledge of the energy consumption in the study region.