K-means based load estimation of domestic smart meter measurements. (15th May 2017)

• Record Type: Journal Article
• Authors: Al-Wakeel, Ali; Wu, Jianzhong; Jenkins, Nick
• Is Part Of: Applied energy. Volume 194(2017)
• Page Start: 333
• Page End: 342
• Publication Date: 2017-05-15
• Subjects: Cluster analysis -- k-means -- Smart meter measurements -- Load estimation; Power (Mechanics) -- Periodicals; Energy conservation -- Periodicals; Energy conversion -- Periodicals; 621.042
• DOI: 10.1016/j.apenergy.2016.06.046
• Languages: English
• ISSNs: 0306-2619

Abstract

Highlights: A load estimation algorithm based on k -means cluster analysis was developed. Canberra, Manhattan, Euclidean, and Pearson correlation distances were investigated. Daily and segmented load profiles of aggregated smart meters were used. Canberra distance outperforms the other distance functions. High accuracy estimates were obtained with cluster centres between 16 and 24 h. Abstract: A load estimation algorithm based on k -means cluster analysis was developed. The algorithm applies cluster centres – of previously clustered load profiles – and distance functions to estimate missing and future measurements. Canberra, Manhattan, Euclidean, and Pearson correlation distances were investigated. Several case studies were implemented using daily and segmented load profiles of aggregated smart meters. Segmented profiles cover a time window that is less than or equal to 24 h. Simulation results show that Canberra distance outperforms the other distance functions. Results also show that the segmented cluster centres produce more accurate load estimates than daily cluster centres. Higher accuracy estimates were obtained with cluster centres in the range of 16–24 h. The developed load estimation algorithm can be integrated with state estimation or other network operational tools to enable better monitoring and control of distribution networks.