A fuzzy reinforcement learning algorithm using a predictor for pursuit-evasion games

Abstract

In a pursuit-evasion game, the pursuer learning its strategy by any learning algorithm usually captures the evader when the environment of the game is similar to the environment that the pursuer was trained on. However, the trained pursuer may not be able to capture the evader if the environment of the pursuit-evasion game is different from the training environment. In this paper, we propose a fuzzy reinforcement learning algorithm so that the ability of the pursuer to capture the evader, in a pursuit-evasion game, will increase even when the environment of the game is different from the training environment. The proposed algorithm predicts the future position of the evader using a Kalman filter and then tunes the fuzzy logic controller (FLC) of the pursuer so that the pursuer moves directly to the expected position of the evader, where the capture of the evader will occur. The proposed algorithm is called the Kalman filter fuzzy actor critic learning (KFFACL) algorithm. The proposed KFFACL algorithm is applied to pursuitevasion games that have environments different from the training environment. Simulation results show that the proposed KFFACL algorithm outperforms the state-of-the-art fuzzy reinforcement learning algorithms in terms of the ability of the pursuer to capture the evader and the capture time.