文章基本信息

标题：Multi-agent management of integrated food-energy-water systems using stochastic games: from Nash equilibrium to the social optimum
本地全文：下载
作者：Milad Memarzadeh ; Scott Moura ; Arpad Horvath 等
期刊名称：Environmental Research Letters
印刷版ISSN：1748-9326
电子版ISSN：1748-9326
出版年度：2020
卷号：15
期号：9
页码：940a4
DOI：10.1088/1748-9326/abadca
出版社：IOP Publishing Ltd
摘要：System-level integration and optimization of food-energy-water systems (FEWS) require coordination of multiple agencies and decision-makers and incorporating their interdependence. In general, such coordination might be hard to achieve. As a result, the literature on FEWS management either optimizes the operations for one sector (or one decision-maker), or models interdependence among the sectors without optimizing their operations. In this article, we develop a novel multi-agent management optimization approach that is able to incorporate stochasticity and uncertainty in the system's dynamics and interdependence of the water and energy resources for food production. The proposed method is the first attempt to utilize fundamentals of decision and game theories to optimize operations of multi-agent FEWS. We specifically focus on differentiating between (1) cooperative decision optimization of the operations, where all decision-makers cooperate to achieve the best outcome for the whole system, the social optimum, and (2) non-cooperative decision-making of the agents, the Nash equilibrium. Illustrating with a real-world case study of FEWS in Ventura County, California, we show the difference between the cooperative and non-cooperative decision making in terms of long-term expected cost of managing the system. We further show how the extra costs associated with utilizing the renewable sources of water and energy could be incentivised, so that the non-cooperative solution (the Nash equilibrium) would naturally converge to the best outcome for the whole system (the social optimum).
关键词：Food-energy-water systems; Game theory; Multi-agent management