Learning Through Fictitious Play in a Game-Theoretic Model of Natural Resource Consumption

Manzoor, T., Rovenskaya, E. ORCID: https://orcid.org/0000-0002-2761-3443, Davydov, A., & Muhammad, A. (2018). Learning Through Fictitious Play in a Game-Theoretic Model of Natural Resource Consumption. IEEE Control Systems Letters 2 (1) 163-168. 10.1109/LCSYS.2017.2777898.

Preview

Text mrmd_lcss_2017.pdf - Accepted Version Available under License Creative Commons Attribution. Download (527kB) | Preview

Abstract

Understanding the emergence of sustainable behavior in dynamic models of resource consumption is essential for control of coupled human and natural systems. In this letter, we analyze a mathematical model of resource exploitation recently reported by the authors. The model incorporates the cognitive decision-making process of consumers and has previously been studied in a game-theoretic context as a static two-player game. In this letter, we extend the analysis by allowing the agents to adapt their psychological characteristics according to simple best-response learning dynamics. We show that, under the selected learning scheme, the Nash Equilibrium is reachable provided certain conditions on the psychological attributes of the consumers are fulfilled. Moreover, the equilibrium solution obtained is found to be sustainable in the sense that no players exhibit free-riding behavior, a phenomenon which occurs in the original open-loop system. In the process, via a Lyapunov-function based approach, we also provide a proof for the asymptotic global stability of the original system which was previously known to be only locally stable.

Actions (login required)

View Item