Distributed Demand Response for Multienergy Residential Communities With Incomplete Information
Weifeng Zhong,Kan Xie,Yi Liu,Chao Yang,Shengli Xie,Yan Zhang
DOI: https://doi.org/10.1109/tii.2020.2973008
IF: 12.3
2021-01-01
IEEE Transactions on Industrial Informatics
Abstract:This article proposes distributed demand response (DR) approaches for a multienergy residential community, which is equipped with various energy conversion and storage devices to serve multiple residential loads (e.g., electricity, natural gas, and heating loads). In the proposed DR approaches, each of the energy devices and loads is an individual decision-maker and also a node in a randomly connected communication network. The DR approaches are tolerant to incomplete information which is caused by random inaction of nodes and links in the network. At first, in order to coordinate nodes' behaviors in distributed DR, different information transmission mechanisms among nodes are employed. Particularly, Steiner tree broadcast, in which nodes are networked according to their energy types, is proposed to lower the nodes' computational complexity and the network's communication overhead. Based on the information transmission mechanisms, the initial DR problem is transformed into network problems that are solvable in a random network. Then, based on the randomized alternating direction method of multipliers, distributed algorithms are designed to optimally solve the network problems in the presence of incomplete information. In simulation, real-world datasets of multiple energy loads and prices are used, and three proposed DR approaches are compared in terms of convergence performance and communication overhead.
automation & control systems,computer science, interdisciplinary applications,engineering, industrial