Comparison of K-Means and Hierarchical Clustering Used in Power Consumption Analysis

In order to analysis the characteristics of the electricity consumption, a typical consumer’s electricity consumption data from January 1st 2022 to July 27th 2022 is collected. K-Means and hierarchical clustering models are built. The highest temperature, the lowest temperature and the consumption are set as input of the model. The results show that when the cluster of K-Means is 3, although the results can reflect the heating demand, but can not reflect the refrigeration demand. The results of hierarchical clustering can reflect the heating and refrigeration demand both when group average and furthest neighbor is used. The results of hierarchical clustering is more credible compared with K-Means.