A Scalable and Distributed Algorithm for Managing Residential Demand Response Programs Using Alternating Direction Method of Multipliers (ADMM)

• Published in: IEEE transactions on smart grid Vol. 11; no. 6; pp. 4871 - 4882
• Main Authors: Kou, Xiao, Li, Fangxing, Dong, Jin, Starke, Michael, Munk, Jeffrey, Xue, Yaosuo, Olama, Mohammed, Zandi, Helia
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.11.2020; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Alternating direction method of multipliers (ADMM); Convex functions; Customers; Demand side management; demand-side resources (DSR); Distribution networks; Electric power demand; home energy management systems (HEMS); HVAC; Load management; Load modeling; Multipliers; Optimization; Peak load; POWER TRANSMISSION AND DISTRIBUTION; Privacy; residential demand response management systems; Temperature; Water heating
• ISSN: 1949-3053; 1949-3061
• DOI: 10.1109/TSG.2020.2995923

For effective engagement of residential demand-side resources and to ensure efficient operation of distribution networks, we must overcome the challenges of controlling and coordinating residential components and devices at scale. In this paper, we present a distributed and scalable algorithm with a three-level hierarchical information exchange architecture for managing the residential demand response programs. First, a centralized optimization model is formulated to maximize community social welfare. Then, this centralized model is solved in a distributed manner with alternating direction method of multipliers (ADMM) by decomposing the original problem to utility-level and house-level problems. The information exchange between the different layers is limited to the primary residual (i.e., supply-demand mismatch), Lagrangian multipliers, and the total load of each house to protect each customer's privacy. Simulation studies are performed on the IEEE 33 bus test system with 605 residential customers. The results demonstrate that the proposed approach can reduce customers' electricity bills and reduce the peak load at the utility level without much affecting customers' comfort and privacy. Finally, a quantitative comparison of the distributed and centralized algorithms shows the scalability advantage of the proposed ADMM-based approach, and it gives benchmarking results with achievable value for future research works.