Task Decision-Making and Obstacle Clearing Optimization of Police Agents in RoboCup Rescue Simulation System

In recent years, there have been various disasters caused by both natural phenomena and human activities, threatening the life and safety of human beings. Therefore, disaster rescue has become an urgent research topic and has attracted a great deal of attention from researchers. With the repaid development of robotics, sensor technology and wireless communication, multi-agent cooperative systems has been widely used in disaster rescue scenarios such as earthquakes to improve rescue efficiency. In this paper, RoboCup Rescue simulation system is employed as a research platform, and the target selection and obstacle clearing task optimization of police agents in disaster scenarios has been studied. By utilizing reinforcement learning methods, the clearing strategy of police agents is optimized, enhancing task execution efficiency and agent collaboration. An optimization scheme based on muti-agent is proposed, in which a reasonable task prioritization mechanism, cooperation strategy and clearing policy are designed. A reinforcement learning framework is constructed to optimize clearance actions, and simulation experiments are conducted to verify the effectiveness of the proposed optimization scheme in various disaster scenarios. Experimental results demonstrate that the optimized strategy significantly improves the execution efficiency of clearance tasks, providing theoretical foundations and practical references for future intelligent rescue systems.