Abstract: This article presents a safe reinforcement learning algorithm for the zero-sum game (ZSG) problem of hypersonic flight vehicles within the actor-critic-disturbance framework. Initially, a ...