Multi-Robot Cooperation Strategy in Game Environment Using Deep Reinforcement Learning

Abstract

The multi-robot system combines the characteristics and advantages of each component robot and can break through the constraints of a single robot's capability, greatly expanding the application of the robot. However, in the game environment, multi-robot systems face the challenge of intelligent decision-making in high-dimensional complex dynamic environments. The research progress of multi-agent decision-making strategies in the game environment based on deep reinforcement learning provides a solution for solving the problems faced by multi-robot systems. To this end, based on the deep reinforcement learning method, we analyze the multi-agent collaboration strategy in the game environment and propose a learning method that can measure cooperative information between multiple agents. On this basis, we conduct a Nash equilibrium game strategy analysis on the specific multi-agent game problem-the territory defense, use deep Q learning method to learn the defender's joint defense strategy. We conducted simulation experiments and verified the effectiveness of our method. Furthermore, we conducted experiments on the actual multi-robot system platform and demonstrated the feasibility of multi-agent cooperation strategy in practical multi-robot system based on deep reinforcement learning.