Research Article
Power Consumption Behavior Analysis Method Based on Improved Clustering Algorithm of Big Data Technology
@INPROCEEDINGS{10.1007/978-3-031-50577-5_23, author={Zheng Zhu and Haibin Chen and Shuang Xiao and Jingrui Yan and Lei Wu}, title={Power Consumption Behavior Analysis Method Based on Improved Clustering Algorithm of Big Data Technology}, proceedings={Multimedia Technology and Enhanced Learning. 5th EAI International Conference, ICMTEL 2023, Leicester, UK, April 28-29, 2023, Proceedings, Part III}, proceedings_a={ICMTEL PART 3}, year={2024}, month={2}, keywords={Big Data Technology Power Consumption Data Power Consumption Behavior of Users Improvement of Clustering Algorithm Behavior Analysis}, doi={10.1007/978-3-031-50577-5_23} }- Zheng Zhu
Haibin Chen
Shuang Xiao
Jingrui Yan
Lei Wu
Year: 2024
Power Consumption Behavior Analysis Method Based on Improved Clustering Algorithm of Big Data Technology
ICMTEL PART 3
Springer
DOI: 10.1007/978-3-031-50577-5_23
Abstract
With the increase of the types and quantities of electrical equipment, the magnitude of user power consumption data has increased exponentially. Deep mining and analysis of it is the key to help the power grid understand customer needs. Therefore, the research on user power consumption analysis method based on improved clustering algorithm based on big data technology is proposed. Analyze the influencing factors of users’ electricity use behavior (economic factors, time factors, climate factors and other factors), on this basis, explore the characteristics of users’ electricity load, reduce and process users’ electricity data samples based on PCA algorithm, improve clustering algorithm by using big data technology – extreme learning machine principle, cluster and process users’ electricity use behavior, use GSP algorithm to mine sequential patterns of users’ electricity use behavior, and obtain users’ electricity use rule model, Thus, the analysis of users’ electricity consumption behavior is realized. The experimental data show that the minimum value of MIA index obtained by applying the proposed method is 0.09, and the maximum accuracy of the user power consumption law model is 92%, which fully confirms that the proposed method has better application performance.
