Adaptive Dynamical Programming for Multiplayer Zero-Sum Games of Continuous-Time Markov Jump Linear Systems
Huan Chen, Xilin Xin, Shuping He · 2022 37th Youth Academic Annual Conference of Chinese Association of Automation (YAC) · 2022
This work solve the problem of multiplayer zero-sum games (MZSGs) by using an adaptive dynamic programming (ADP) algorithm. First, the problem is considered as an equivalent problem of solving the coupled game algebraic Riccati equations (CGARE). Then, in response to the complexity of the Markov jump systems (MJSs), a novel conversion technology is introduced to solve the coupling relationship between subsystems of the MJSs and design an online policy algorithm. Meanwhile, the probing noise is added to stimulate the system. Finally, simulations involving a multiplayer case shows the correctness of the designed control algorithm.