Joint Search And Rescue At Sea Based On Swarm Intelligence Algorithm

Maritime trade and transportation activities bear most of the tasks of world commodity circulation at this stage,and play an important role in the world commodity trade.The bad marine working environment increases the occurrence of marine accidents,which makes marine search and rescue become an important research topic.At present,the development of search and rescue work mainly depends on human resources.In this process,it is necessary to make decisions manually,collect and analyze intelligence,further distribute search and rescue tasks,assign search and rescue areas and arrange ships to carry out search and rescue.At the same time,a,large number of personnel and ship equipment need to be put into the field search for the drift caused by the unclear coordinates of the shipwreck location and the influence of current,wind and other factors.The efficiency is slow.For the most dependent communication technology in the search and rescue process,after the search and location of the crash area is completed,the offshore area can carry out rescue command through the shore based cellular base station or radio communication.However,in the far sea area far away from the shore base,there is a lack of communication means,and only a few satellite resources can be used,so the rescue command work on the search and rescue scene becomes difficult.With the development of unmanned ship vehicle(USV)and unmanned aerial vehicle(UAV),its distributed,cooperative,parallel robustness and intelligence provide a new idea for the development of maritime search and rescue.In this paper,based on the needs of multi-agent search and rescue process,we build a multi-agent search and rescue model at sea,and build a hierarchical control network system composed of multi-agent.The intelligent individuals in the system rely on the distributed parallel way to carry out edge computing and group search and rescue.Through information sharing and collaborative decision-making,the search and rescue task is completed among the search and rescue intelligent groups.The whole system consists of base station,satellite and multi-agent group.The multi-agent component is divided into three layers,which are central control agent,group control agent and search and rescue agent.The commander on the shore sends scheduling commands to the multi-agent group through the base station and satellite.The central control agent is responsible for the location of the search and rescue sea area and the sending of search and rescue tasks.Group control agent is responsible for shortest route planning and search path planning.Search and rescue agents are responsible for specific search and rescue work.We use ant colony algorithm to complete the path planning task in the system,and improve the ant colony algorithm to improve the convergence speed of the algorithm.At the same time,in the process of rescue,we use particle swarm optimization algorithm to schedule data packets intelligently and optimize the performance of network forwarding.In the system designed in this paper,the distributed search and rescue task allocation scheme breaks through the limitations of task allocation relying on manpower in the past.The edge computing function of each group of control agents reduces the network communication load,and improves the system operation efficiency through independent decision-making,key information sharing and task collaboration.Finally,we verify the performance of the improved algorithm and realize the system functions through simulation.