Algorithms for zero-sum stochastic games with the risk-sensitive average criterion
Optimization and Control
2025-05-08 v1
Abstract
This paper is an attempt to compute the value and saddle points of zero-sum risk-sensitive average stochastic games. For the average games with finite states and actions, we first introduce the so-called irreducibility coefficient and then establish its equivalence to the irreducibility condition. Using this equivalence,we develop an iteration algorithm to compute -approximations of the value (for any given ) and show its convergence. Based on -approximations of the value and the irreducibility coefficient, we further propose another iteration algorithm, which is proved to obtain -saddle points in finite steps. Finally, a numerical example of energy management in smart grids is provided to illustrate our results.
Cite
@article{arxiv.2505.04546,
title = {Algorithms for zero-sum stochastic games with the risk-sensitive average criterion},
author = {Fang Chen and Xianping Guo and Xin Guo and Junyu Zhang},
journal= {arXiv preprint arXiv:2505.04546},
year = {2025}
}