Differential Game-Theoretic Analysis on Information Availability in Decentralized Demand-Side Energy Management Systems

Ryohei ARAI  Koji YAMAMOTO  Takayuki NISHIO  Masahiro MORIKURA  

Publication
IEICE TRANSACTIONS on Communications   Vol.E97-B   No.9   pp.1817-1825
Publication Date: 2014/09/01
Online ISSN: 1745-1345
DOI: 10.1587/transcom.E97.B.1817
Type of Manuscript: Special Section PAPER (Special Section on Ambient Intelligence and Sensor Networks)
Category: 
Keyword: 
demand-side energy management,  smart grids,  differential game,  optimal control,  information availability,  decentralized control,  

Full Text: PDF>>
Buy this Article




Summary: 
Differential games are considered an extension of optimal control problems, which are used to formulate centralized control problems in smart grids. Optimal control theory is used to study systems consisting of one agent with one objective, whereas differential games are used to formulate systems consisting of multiple agents with multiple objectives. Therefore, a differential-game-theoretic approach is appropriate for formulating decentralized demand-side energy management systems where there are multiple decision-making entities interacting with each other. Moreover, in many smart grid applications, we need to obtain information for control via communication systems. To formulate the influence of communication availability, differential game theory is also promising because the availability of communication is considered as part of an information structure (i.e., feedback or open-loop) in differential games. The feedback information structure is adopted when information for control can be obtained, whereas the open-loop information structure is applied when the information cannot be obtained because of communication failure. This paper proposes a comprehensive framework for evaluating the performance of demand-side actors in a demand-side management system using each control scheme according to both communication availability and sampling frequency. Numerical analysis shows that the proposed comprehensive framework allows for an analysis of trade-off for decentralized and centralized control schemes.