Minimax Equilibria of Discrete-Time Hybrid Markov Games
Open Access
- 21 August 2025
- journal article
- research article
- Published by Springer Nature in Dynamic Games and Applications
- Vol. 16 (4) , 1-36
Abstract
This paper studies a discrete-time dynamic stochastic game featuring two players: a Leader (Player 1) and a Follower (Player 2), distinguished by its hybrid nature. Specifically, the system’s dynamics adhere to a Markovian-like property, further divided into two sub-dynamics: one regular and the other impulsive, both exerting influence on state evolution. Moreover, players are equipped with two action types: regular and impulsive. The activation of each sub-dynamics depends on the chosen action type. The follower seeks a strategy, conceptualized as a sequence of actions, aimed at maximizing a specified infinite-horizon discounted payoff criterion, where the discount factor is contingent upon the history of state-action pairs. Conversely, the leader aims to minimize such a payoff through its strategy, giving rise to the concept of a minimax equilibrium. Our analysis hinges upon the dynamic programming method, which enables us to (i) characterize the optimal value function as a solution to certain dynamic programming equations and (ii) establish the existence of an optimal minimax equilibrium. We conclude by presenting the application of our findings through a pollution management scenario involving multiple contingency levels often encountered by foreign industries or companies (followers) operating within a particular country (leader).Keywords
Funding Information
- Secretaría de Ciencia, Humanidades, Tecnología e Innovación (0)
- Secretaría de Ciencia, Humanidades, Tecnología e Innovación (87787)
This publication has 22 references indexed in Scilit:
- Discrete-time control with non-constant discount factorMathematical Methods of Operations Research, 2020
- Discrete-Time Hybrid Control in Borel SpacesApplied Mathematics & Optimization, 2018
- Approximation of Discounted Minimax Markov Control Problems and Zero-Sum Markov Games Using Hausdorff and Wasserstein DistancesDynamic Games and Applications, 2018
- Stochastic Games with Unbounded Payoffs: Applications to Robust Control in EconomicsDynamic Games and Applications, 2011
- Verification of discrete time stochastic hybrid systems: A stochastic reach-avoid decision problemAutomatica, 2010
- Probabilistic reachability and safety for controlled discrete time stochastic hybrid systemsAutomatica, 2008
- Reachability Analysis for Controlled Discrete Time Stochastic Hybrid SystemsLecture Notes in Computer Science, 2006
- Stochastic Hybrid ControlJournal of Mathematical Analysis and Applications, 2000
- Further Topics on Discrete-Time Markov Control Processes1999
- Discrete-Time Markov Control Processes1996