A zero-sum hybrid stochastic differential game with switching controls

Siyu Lv,Jie Xiong
DOI: https://doi.org/10.3934/mcrf.2024007
2024-03-14
Mathematical Control and Related Fields
Abstract:In this paper, we study a zero-sum stochastic differential game in an infinite horizon, where the state equation consists of a number of diffusions coupled by a Markov chain and both players in the game employ switching controls. Based on the dynamic programming principle (DPP), the lower and upper value functions of the game are characterized as the unique viscosity solution of the associated Hamilton-Jacobi-Bellman-Isaacs (HJBI) equation, which turns out to be two systems of variational inequalities with bilaterally inner-connected obstacles; thus the two value functions coincide and the game admits a value. Moreover, a verification theorem as a sufficient condition for Nash equilibriums is established, in which the equilibrium switching strategies for the two players are given in terms of the obstacle part of the HJBI equation.
mathematics, applied
What problem does this paper attempt to address?