Title: Power load clustering algorithm for demand response

Authors: Yanguang Cai; Helie Huang; Hao Cai; Yuanhang Qi

Addresses: School of Automation, Guangdong University of Technology, Guangzhou, Guangdong, 510006, China ' School of Automation, Guangdong University of Technology, Guangzhou, Guangdong, 510006, China ' School of Automation, Guangdong University of Technology, Guangzhou, Guangdong, 510006, China ' School of Automation, Guangdong University of Technology, Guangzhou, Guangdong, 510006, China

Abstract: Satisfactory clustering of power load is an essential prerequisite for the effective implementation of demand response (DR) programs. Focusing on the inability of common clustering algorithms to specify the similarity degree between load profiles; this paper proposes a novel power load similarity measurement criterion based on the maximum deviation, similarity degree and deviation degree, termed maximum deviation similarity criterion (MDSC). We further propose a power load clustering algorithm based on the MDSC for obtaining reasonable load classification. The proposed MDSC is capable of specifying the similarity degree and effectively describes the shape similarity between load profiles. Furthermore, the criterion is simple, reasonable and flexible in nature. A case study with 32 load data clustering analysis is used to verify the proposed clustering algorithm. Experimental results demonstrate that the proposed clustering algorithm is computationally faster and has a better clustering efficiency, allowing it to better meet the needs of DR programs.

Keywords: demand response; power load; maximum deviation similarity criterion; MDSC; clustering algorithm; shape similarity.

DOI: 10.1504/IJAACS.2019.096662

International Journal of Autonomous and Adaptive Communications Systems, 2019 Vol.12 No.1, pp.34 - 49

Available online: 30 Oct 2018 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article