Research on Performance Evaluation of Wargame System Based on Deep Reinforcement Learning

Abstract

Abstract Deep reinforcement learning combines the advantages of deep learning and reinforcement learning to make end-to-end perception decisions in complex high-dimensional state action spaces. The paper proposes a deep reinforcement learning method for the deduction of the wargame system, which can better assist the combat commander in wartime decision-making. The simulation proves the effectiveness of the method.