Research On Task Allocation Method Of Heterogeneous Multi-UAVs For Area Detection

With the gradual development of UAV technology,the application of drones in the field of detection has become more and more in-depth.At the same time,with the increasing complexity of detection,how to utilize the limited UAV resources to complete the detection mission has become an urgent problem.Therefore,in order to accomplish the detection mission better,it is of great practical significance to study the task allocation problem of the UAV in the regional detection mission.The problem of task allocation for area detection with heterogeneous multi-UAVs is how to allocate a series of target area for each drone as the detection task,and at the same time plan to detect the execution sequence and flight path of these target areas so as to optimize the total length of the flight and the total detection profits simultaneously.So the task allocation problem in this thesis is a multi-objective optimization problem.At the same time,unlike the traditional point-target problem,the regional target needs to consider both the flight path between areas and the intra-area detection path.The joint planning of the two paths can achieve the optimization of the flight path,and only the joint planning of the two paths can achieve the optimization of the flight path.In this thesis,the multi-objective model of Dubins Heterogeneous Vehicle Routing Problem(DHVRP)is established to solve the task allocation of heterogeneous UAV regional detection mission considering the Dubins path constraints of UAVs,sensor storage constraints,sensor detection performance and multi-typed region detection tasks.The non-dominated sorting genetic algorithm(NSGA-II)algorithm based on Pareto optimal solution is used in this thesis,and the chromosome coding,crossover operation and mutation operation of the algorithm are designed in a targeted way to solve the task allocation problem.The results of simulation experiments show that the NSGA-II algorithm used in this thesis can effectively solve the task allocation problem.The overall result of the NSGAII solution is superior to the genetic algorithm under the single objective and artificial task assignment planning method.The research results of this thesis are helpful to use the limited UAV and sensor resources more effectively in the detection task,and have significance to solve the task allocation problem of the UAV.