Clustering of Electricity Usage Using K-Means Method

The electricity usage in the buildings depends on all activities related to using any electrical devices in the buildings. More electrical devices in the building used, the more electrical energy will be used as well. Using excessive electrical energy can cause the waste of electrical energy in buildings. The waste of electrical energy also can happen at the campus. The campus has some needs for electricity energy in working hours even outside working hours. As a result, the logistics team can not determine the regulation of using electrical energy on campus and can causes the overuse of electricity and the swelling of the electricity bill. Therefore, to solve this problem it needs a system that can clustering data that results in excessive power usage or not. Then the logistics team can do something about how to prevent the overuse of electricity and the swelling of the electricity bill. This paper will be using K-Means Clustering to cluster data regarding how long the time it takes that using any electricity in the building. The results clustered data then will be searched for the accuracy value using the Davies Boulden Index method. The accuracy that has been obtained is 0,83 and the data that has been clustered into 3, which is low usage, medium usage, and high usage. This paper will focus on making the K-Means clustering system required by theuser.