Research On Task Assignment Of Multiple UAVs

With the continuous development of modern information technology and UAV countermeasure technology,as a new military weapon,UAV has an increasingly broad application prospect in modern war with its excellent performance and relatively low cost.In order to meet the requirements of the actual battlefield environment and mission diversity,as well as improve the task execution efficiency and task completion rate of UAVs,the theory of multi-UAVs task allocation emerged.Through the research and analysis of the existing UAV task allocation theory,in view of the current task allocation model is not clear about the characteristics of task execution and the low efficiency problem when solving large-scale problems,this paper mainly studies from the following aspects:1.By analyzing the scenarios of UAV reconnaissance and attack,the characteristics of these two kinds of tasks are summarized.At the same time,aiming at the problem of target value evaluation of strike mission,this paper introduces the goal intention value into the goal value evaluation scheme.In the traditional target value evaluation scheme,only the objective attributes of the target,such as speed,distance and height,are usually considered.In this paper,the intended value of the target is quantitatively described from the coupling of information such as route shortcut,height and distance,and it is weighted with the objective attribute value to get the final value of the target.2.According to the characteristics of a large number of tasks in reconnaissance task assignment,which is related to the execution sequence,this paper proposes a reconnaissance task assignment model based on the MTSP and considering the survival probability of UAV.In order to solve the problem of reconnaissance task allocation,this paper proposes the improved genetic algorithm based on group local search strategy.The algorithm is based on the classical genetic algorithm,and uses group selection strategy and 3-Opt-like local search strategy to improve the convergence speed and global search ability of the algorithm.Finally,a comparative experiment proves that the proposed algorithm is suitable for solving the problem of reconnaissance assignment.3.According to the characteristics of large number of UAVs,failure probability and weak correlation with strike sequence in strike assignment,based on the WTA model,the concept of weapon platform is abandoned,the average range cost of UAVs is introduced as the second objective function,and a multi-objective version of strike task allocation model is proposed.Since abandoning the weapon platform will increase the scale of task assignment problem,this paper proposes a non-dominant artificial bee colony algorithm based on dual adaptive strategy to solve the problem of low efficiency in solving the task assignment problem of large-scale strikes.The algorithm solves the problem of multiple objective functions by using nondominant sorting strategy,improves the convergence speed by using two adaptive strategies,and guarantees the global search performance by Pareto guidance strategy.Finally,the effectiveness of the two adaptive strategies and the effectiveness of the proposed algorithm in solving the task assignment problem of large-scale attack are proved by several sets of experiments.