查看文章

Higher education institutional building clusters (IBCs) have a considerable socioeconomic impact on the local community. Thus, optimal energy management in IBCs offers a significant opportunity to enhance their energy efficiency at a community level. However, for implementing energy management strategies in IBCs, it is essential to analyze the electrical demand profiles of the buildings to comprehend their time-varying operational behaviour. This study introduces a novel cloud-oriented quantile-based data mining (DM) framework that investigates the energy demand profiles of buildings in IBC and draws a cross-building comparison regarding their energy use dynamics and quantifies their electrical impact during peak demand scenarios. The proposed DM framework not only analyses the abnormal electrical demand patterns but also identifies the coincident peaking (CP) behaviour of buildings in regard to IBC’s energy demand. Further, to establish a cross-building energy demand comparison, buildings are linearly ranked considering two major criteria (i) identifying buildings that exhibit frequent fluctuations in their electrical demand,(ii) identifying buildings that possess a high electrical impact of cluster-wide electrical demand. Lastly, an R-shiny-based dashboard is also developed to provide IBC managers with a visually intuitive display of the DM framework. In addition, the practical evaluation of the DM framework is accomplished by investigating the electrical demand profiles of the buildings at The University of British Columbia, Okanagan Campus.