An energy efficiency solution based on time series data mining algorithm - a case study of small hotel

Abstract

This study aims to conduct data mining research on the time series energy consumption dataset of a small hotel. Earlier studies on data mining have demonstrated that cluster and association analysis had been commonly used methods today, while this has not yet been investigated under time series dimension. For that consequence, this article utilizes K-shape and Apriori algorithm coded by Python language to explore the time series energy consumption data of the small hotel in terms of the subentry and total energy consumption. Final results reveal that the energy consumption curve and association rules can effectively reflect the working characteristics of the small hotel. From the clustering results, the small hotel working feature and weather condition determine the time series energy consumption curve shape. As for association rules, there is a different chain relationship between the energy consumption of each subentry and total energy consumption of the small hotel, especially the consumption of heating gas and cooling, which mainly determines the changes in other energy consumption.