An energy efficiency solution based on time series data mining algorithm on elementary school building

Abstract

The purpose of this study is to conduct data mining research on the time series energy consumption dataset of primary school buildings. Researches done so far mainly focused on cluster analysis and association analysis. In this article, python language is used as the carrier; Κ-shape algorithm and apriori algorithm are adapted to perform cluster analysis and association analysis on the time series energy consumption data of primary school buildings, from the perspective of the subentry and total energy consumption. The final result shows that the energy consumption curve and association rules obtained through clustering and association can effectively reflect the operating characteristics of primary school buildings. In terms of clustering, regardless of the subentry energy consumption or the total energy consumption, there are different characteristics of energy consumption patterns in winter and summer, which depend on the school schedule. With respect to association rules, there are different chain relationships between itemized total energy consumptions in the primary school building. In particular, heating natural gas consumption and cooling consumption mainly determine the changes in other energy consumption.