Research On Multi Uav Cooperative Path Planning Based On Swarm Intelligence

With the popularization and development of information and communication technology,as well as the improvement of underlying hardware performance,the concept of "everything connected" has penetrated into all levels of human society,and the field of cluster intelligence is developing rapidly in this wave.As the most common application scenario of cluster intelligence,large-scale unmanned system is expected to be able to transplant cluster intelligence technology to unmanned system,so that it has more powerful efficiency to solve all kinds of complex tasks,and has the ability of independent decision-making to save human cost.For the unmanned system,path planning is often used as a necessary component to support the autonomous action of the cluster.In the context of clustering,the path planning algorithm and its dependent problem model need to be adjusted accordingly.Considering the complexity of the actual environment,we need to build a reasonable model and retain the main characteristics of the environment.On the other hand,when the cluster scale increases,the decision-making cost of the system will accumulate rapidly,so it is necessary to design a reasonable decision-making mechanism to reduce the decision-making cost of the whole system.In addition,for clusters,decentralized collaboration is more meaningful than centralized collaboration,so it is necessary to introduce a reasonable mechanism to make the cluster maintain the efficiency of tasks as much as possible while ensuring the decentralization.In this paper,firstly,the global environment information is modeled discretely in a grid way,and the update and maintenance mechanism of each node situation is designed in detail considering the cluster networking mode.Based on the simple graph search method,the concept of decision switch is introduced to reduce the problem size in time scale and improve the decision consumption of the whole system.On the other hand,in order to ensure the overall efficiency and persistence of the system in the context of large-scale spatial area,this paper also introduces the idea of "multi-level collaborative path planning" and designs the preliminary architecture based on the problem scale sparseness of spatial scale.It reduces the search space by multi-level partition,and combines local planning paths to approximate the global optimal solution.Finally,the feasibility of the above idea is verified by the simulation experiment.