Integral Reinforcement Learning-Based Guidance Law Design Against Maneuvering Target with Constrained Input

This paper proposes a novel real-time optimal guidance law of the missile interception task via integral reinforcement learning (IRL). Particularly, two typical practical scenarios-target maneuvering and constrained control input are considered. By designing an appropriate reward function, the finite-horizon constrained optimal control problem is transformed into an infinite-horizon optimal control problem. Then, by integrating value function approximation into IRL, the infinite-horizon optimal control problem is solved online, avoiding the use of unknown target information. The effectiveness and superiority of the proposed approach are demonstrated through numerical simulations.