Integrated guidance and control for pinpoint mars landing using reinforcement learning

Future Mars missions will require advanced guidance, navigation, and control algorithms for the powered descent phase in order to target specific surface locations and achieve pinpoint accuracy (landing error ellipse < 5m radius). This requires both a navigation system capable of estimating the lander’s state in real-time and a guidance and control system that can map the estimated lander state to body-frame actuator commands. In this paper we present a novel integrated guidance and control algorithm designed by applying the principles of reinforcement learning theory. The key innovation is the use of reinforcement learning to learn a policy mapping the lander’s estimated state directly to actuator commands, with the policy resulting in accurate and fuel efficient trajectories. Specifically, we use proximal policy optimization, a policy gradient method, to learn the policy. We present simulation results demonstrating the guidance and control system’s performance in a 6-DOF simulation environment, and demonstrate robustness to noise and system parameter uncertainty.