Research On Collaborative Task Allocation And Safety Path Planning Of Multiple UAVs

Unmanned aerial vehicles have shifted from military applications to a wider range of civilian applications.With its high maneuverability,scalability,and intelligent characteristics,it has been widely used in target search,area monitoring,agricultural production and other fields,and has achieved outstanding results.The UAV is equipped with a camera to monitor the ground environment,which can provide huge improvements to the target search task,and provide a strong guarantee for maintaining urban public safety and fighting crime.However,the ability of a single UAV to perform tasks is limited by the endurance of the UAV itself,.Therefore,in the face of large-scale or complex tasks,multiple UAVs performing tasks collaboratively will effectively overcome the shortcomings of the UAV itself and improve the efficiency of task execution.The content of this study will mainly start from the following aspects.(1)Propose a multi-UAV collaborative regional monitoring framework.The multi-machine cooperative area coverage model and the formation and networking methods of multiple heterogeneous drones are proposed.In this part,the calculation method of ground target waypoint coordinate information is proposed.The UAVs in the cluster plan the shortest flight path according to the waypoint to carry out their tasks together.(2)A collaborative task allocation scheme based on K-means algorithm is proposed.First,the multi-task assignment problem is transformed into a combined optimization problem with multiple constraints,and the waypoint task is assigned.The purpose is to allow the UAV to perform the task with the lowest voyage cost.(3)Propose a real-time multi-UAV path planning scheme in dynamic environment.Firstly,the UAV safe anti-collision distance model is proposed,and then the real-time planning method of multiple UAV paths in the group based on the artificial potential field method is proposed.When a new waypoint mission is added or the formation structure changes,the flight path can be adjusted in real time.Finally,simulation experiments and real-machine experiments prove the feasibility of the proposed aerial photography coverage model and task assignment algorithm,and comparison experiments with other methods prove the effectiveness of the proposed method.At the end,the shortcomings of this research work and the follow-up research are discussed.