Abstract
In order to help the pursuer find its advantaged control policy in a one-to-one game in space, this paper proposes an innovative pre-trained fuzzy reinforcement learning algorithm, which is conducted in the x, y, and z channels separately. Compared with the previous algorithms applied in ground games, this is the first time reinforcement learning has been introduced to help the pursuer in space optimize its control policy. The known part of the environment is utilized to help the pursuer pre-train its consequent set before learning. An actor-critic framework is built in each moving channel of the pursuer. The consequent set of the pursuer is updated through the gradient descent method in fuzzy inference systems. The numerical experimental results validate the effectiveness of the proposed algorithm in improving the game ability of the pursuer.
Highlights
Tracking space targets is beneficial for orbital garbage removal, recovery of important components, and early warning of space threats [1]
The differential game in space has more complex dynamics; it will be extremely hard to solve without any prior information
In this paper, we propose an innovative pre-trained fuzzy reinforcement learning (PTFRL) algorithm to help the pursuer optimize its control policy through a pre-training process
Summary
Tracking space targets is beneficial for orbital garbage removal, recovery of important components, and early warning of space threats [1]. The differential game in space has more complex dynamics; it will be extremely hard to solve without any prior information To overcome this shortcoming, in this paper, we propose an innovative pre-trained fuzzy reinforcement learning (PTFRL) algorithm to help the pursuer optimize its control policy through a pre-training process. This paper applies a pre-trained fuzzy reinforcement learning algorithm to optimize the control policy of a pursuer, which is used for a one-to-one game in outer space. The main improvements of this paper are as follows: (1) Unlike the previous control laws, which were designed based on the adaptive control theory, for the first time, we utilize the technique of reinforcement learning to help the pursuer track a moving non-cooperative target in space. The structure of this paper is as follows: Section 2 presents the dynamics of the pursuer and the evader; Section 3 discusses the fuzzy inference system and its combination with reinforcement learning for continuous systems; Section 4 applies the pre-trained fuzzy reinforcement learning algorithm for the pursuer; Section 5 simulates the proposed algorithm; Section 6 discusses the experimental results; Section 7 draws the conclusions
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have