Search

Aiming at the attack and defense game problem in the target-missile-defender three-body confrontation scenario, intelligent game strategies based on deep reinforcement learning are proposed, including an attack strategy applicable to attacking missiles and active defense strategy applicable to a target/defender. First, based on the classical three-body adversarial research, the reinforcement learning algorithm is introduced to improve the purposefulness of the algorithm training. The action spaces the reward and punishment conditions of both attack and defense confrontation are considered in the reward function design. Through the analysis of the sign of the action space and design of the reward function in the adversarial form, the combat requirements can be satisfied in both the missile and target/defender training. Then, a curriculum-based deep reinforcement learning algorithm is applied to train the agents and a convergent game strategy is obtained. The simulation results show that the attack strategy of the missile can maneuver according to the battlefield situation and can successfully hit the target after avoiding the defender. The active defense strategy enables the less capable target/defender to achieve the effect similar to a network adversarial attack on the missile agent, shielding targets from attack against missiles with superior maneuverability on the battlefield.

Aimed at the poor performance of guidance algorithms designed based on a linearized model in active defense under large leading angle deviation, both-way and one-way cooperative sliding mode guidance algorithms based on the smooth nonsingular terminal sliding mode for the defense missile are proposed. The relative kinematics and linearized models of the target, the active defense missile, and the attacking missile are established. In the design process, two smooth nonsingular terminal sliding mode surfaces are constructed based on zero-effort miss distance and zero-effort velocity, as well as their integral values. A tracking differentiator is introduced for excessive initial command deviation to meet the overload constraints of the active missile and the target. The sensitivity of guidance law to the estimated time-to-go error is reduced, and the target is allowed to perform an independent evasive maneuver. The effectiveness of the proposed guidance strategy is verified by numerical simulation and compared to the existing guidance strategies, the high accuracy, anti-chattering, and strong robustness of the proposed guidance algorithm are verified.