User Power Consumption Cluster Analysis Based on Cloud Computing and Improved K-means Algorithm
Yan Xing Song, Yun Su, Hongshan Yang · 2019
Power companies store a huge amount of user historical electricity consumption data, which is in a complex, independent, and scattered state. If the historical electricity consumption data can be reasonably excavated, we can cluster users according to different users' electricity consumption, electricity usage habits and characteristics, etc. These are beneficial to both data management and power forecasting. Based on this, the electricity consumption data of some users in a certain city is used as the data source. Based on cloud computing and the usage of improved K-means algorithm, we could analyze its historical electricity consumption, monthly electricity consumption change, peak-to-valley power consumption per day and other data, which is useful to cluster the power consumption of each user. After clustering, we obtain the information about electricity consumption of users of each category, which presents distinctive features, guides users to passive and lagging in energy-saving renovation, and supports intelligent business analysis and decision-making.