Research On Cooperative Task Allocation For Reconnaissance/Combat UAV

Multiple unmanned aerial vehicle(UAV)cooperative task allocation is one of the key technologies to improve the system utility of multi-UAV cooperative combat.However,the multi-UAV coordination is facing severe challenges such as computational complexity,heterogeneity of UAVs,diversity of tasks,collision avoidance and communication delay among UAVs.In this paper,a set of heterogeneous UAVs composed of reconnaissance,munition and combat UAVs are required to perform a search and attack mission.Two different scenarios are considered:(1)multiple UAVs cooperate to attack ground targets under the management of reconnaissance UAV;(2)heterogeneous multi-UAV cooperative search and attack mission.Aiming at the multiUAV cooperative task allocation and the coupled path planning problems,the model formulation,theoretical basis,method design and simulation experiments are carried out.The main work and contributions of this paper are as follows:(1)The cooperative control model of multiple UAVs is established under the typical battlefield scenario,two typical mission scenarios of heterogeneous multi-UAV coordination are analyzed,a centralized control architecture is designed for multiple UAVs cooperate to attack ground targets under the management of reconnaissance UAV,a distributed control architecture is designed for heterogeneous multi-UAV cooperative search and attack mission.The model of each element involved in the multi-UAV coordination system is established,and the coupling of the task allocation and path planning sub-problem is analyzed from the general model of the multiUAV task planning problem.A novel task allocation model is established considering the task sequence of UAVs and the resource requirement of targets,which solves the problem of poor applicability caused by the fixed allocation relationship between UAV and target in the traditional task allocation model.A decoupling method for the task planning problem is proposed based on the hierarchical control strategy.(2)Aiming at the problem of multiple UAVs simultaneous arrival path planning when there are various single UAV performance constraints and multi-UAV spatial and temporal cooperative constraints,a distributed particle swarm optimization path planning algorithm is proposed based on multi-population coordination.The quantic Pythagorean Hodograph(PH)curve is employed as the path for each UAV,as the curvature of the quantic PH curve is continuous.The interaction mode and the updating method of particle fitness are designed to satisfy the simultaneous arrival constraint and collision avoidance between UAVs.The proposed path planning method can optimize the velocity and path of each UAV at the same time.Simulation results show that the method can generate feasible paths and velocities for multi-UAV that satisfy various constraints and has good convergence.(3)To solve the task allocation problem of multi-UAV cooperative ground attacking mission under the management of reconnaissance UAV,considering the resource requirement of targets,a multi-UAV task allocation method is proposed based on modified genetic algorithm.Firstly,an integrated task allocation and path planning model is established based on the directed graph and discretization of heading angle.In the proposed task allocation method,the target execution sequence is determined according to the situation awareness of the mission aera,which solves the problem of task deadlock caused by the overlapping coalitions.The chromosome crossover and mutation operators are designed to satisfy the constraints in this mission.The proposed method realizes the unified coding and solving of the integrated task allocation and path planning problem.Comparative experiments verity the effectiveness of the method.(4)For the heterogeneous multi-UAV cooperative search and attack mission,to solve the local task reallocation problem under the constraints of communication distance and time delay,a distributed multi-UAV task reallocation method is proposed based on the contract net protocol.Firstly,an optimization model of the local task reallocation when heterogeneous multi-UAV discover new target is established,an information consensus algorithm is designed for the local UAV communication network to solve the conflict resolution problem in the task announcement process.A coalition formation and resource management method are proposed to solve the coalition formation problem considering the resource requirement of targets.This method enables the UAV coalition to obtain the maximum system utility for the pop-up target and the UAVs in the coalition can consume resource in a more balanced way.Simulation results verify the effectiveness of the task reallocation method.