Linear quadratic zero‐sum game for time‐delayed uncertain stochastic systems
Xin Chen,Yue Yuan,Dongmei Yuan,Yu Shao
DOI: https://doi.org/10.1002/oca.3123
2024-03-29
Optimal Control Applications and Methods
Abstract:This study introduces an approach to solve linear quadratic zero‐sum games withintime‐delayed uncertain stochastic systems. Through transforming these games into equivalent uncertain stochastic zero‐sum games without time delay, we derive equilibrium equations. Analytic expressions for saddle‐point equilibrium are obtained, providing insights into optimal strategies. We illustrate our approach with numerical examples and a counter‐terrorism economic game. Furthermore, Table 1 compares ourwork with existing zero‐sum games in various environments. Summary This study focuses on the analysis of a linear quadratic zero‐sum game (LQZSG) for time‐delayed uncertain stochastic systems. To begin, we introduce a general time‐delayed zero‐sum game. Employing an algebraic transformation method, we transform the time‐delayed zero‐sum game into an equivalent uncertain random zero‐sum game without time delay. Subsequently, we present equilibrium equations that streamline the transformation of the uncertain random zero‐sum game into problems solvable as deterministic difference equations. We then investigate the LQZSG involving a linear time‐delayed uncertain stochastic system with a quadratic objective function. Within this framework, we provide a unified framework for solving this type of game and obtaining the analytic expression for the saddle‐point equilibrium of the LQZSG. Additionally, we present a numerical example and a counter‐terrorism economic game to illustrate the applicability of our findings.
automation & control systems,operations research & management science,mathematics, applied