Development And Research Of Multi-agent Rescue Strategy In RoboCup Rescue Simulation System

In recent years,various emergencies and human sabotage activities have led to frequent disasters,threatening the life and safety of all mankind,and arousing the attention of most people.RoboCup Rescue simulation system is a system that first simulates the urban environment of natural disasters such as earthquakes or fires,and then uses robots to carry out rescue work in this scene.In this system,the robot needs to complete tasks such as rescuing the injured in the disaster and extinguishing possible fires,.The purpose is to reduce the loss in the disaster and improve the overall effect of rescue.The robot’s action strategy where in RoboCup Rescue simulation system provides theoretical support and decision-making support for rescuers in rescue activities.With the help of this simulation system,this topic focuses on researching of multi-agent rescue strategies.The main work is as follows:(1)Clear obstacles of Police ForceThe main task of Police Force is to clear obstacles on the road.This paper adopts the cleaning method of the road guide line mechanism to reduce the possibility of the mobile agent being trapped by obstacles.According to whether there is a central agent and the communication status,a clear obstacles of Police Force based on task priority and Hungarian algorithm is proposed.The effectiveness of the obstacle clearing strategy is confirmed in the simulation experiment.(2)Fire extinguishing strategy of Fire BrigadeThe main task of Fire Brigade is to extinguish and control the fire.This article adopts a zoned fire extinguishing strategy.Use K-means++ clustering algorithm to create map partitions and divide the simulation map into several partitions.Based on the partition,the fire protection agent is adjusted and distributed,and the Graham algorithm is used to establish a convex hull for the burning buildings in each partition.Fire Brigade then controls the fire of the building in the convex hull to prevent the fire from spreading.The effectiveness of the fire extinguishing strategy is confirmed in simulation experiments.(3)Rescue strategy of Ambulance TeamThe main task of Ambulance Team is to rescue the injured agent.The A* algorithm is used to realize the path planning of the mobile agent.In order to quickly collect the list of the wounded,a self-perceived search strategy for the rescue agent and a collaborative search strategy for the firefighting agent are proposed.In order to rescue more wounded,make the rescue agent choose the best target for rescue,adopt a rescue strategy based on task priority,choose according to the rescue target weight,the health of the injured agent and the distance from the rescue agent to the target.The effectiveness of the rescue strategy is confirmed in simulation experiments...