An Exoatmospheric Homing Guidance Law Based on Deep Q Network

Abstract

A homing guidance law for exoatmospheric interceptor based on the Deep Q Network (DQN) algorithm is proposed in this paper. Aiming at the exoatmospheric interception problem, the guidance agent is built with the help of the deep reinforcement learning theory, and the action command is given according to the measurement information of the exoatmospheric interceptor for the accurate interception of the target. The homing guidance problem is first transformed into a Markov decision process, and a three-dimensional (3D) interception scenario is established. Then, the reward function considering the line-of-sight (LOS) rate and the final zero-effort-miss (ZEM) is designed, and the homing guidance problem is transferred to the reinforcement learning framework. After that, DQN is utilized to solve the exoatmospheric interception problem, and the guidance agent is obtained through a large amount of training. Finally, the guidance performance of DQN homing guidance law is verified by numerical simulation examples and compared with the classical true proportional navigation (TPN) guidance law. The results show that the guidance performance of the homing guidance law is better than that of TPN.