A Study Of Online Collaborative Task Allocation Technology For Multiple Unmanned Combat Aerial Vehicles

Online collaborative task allocation for multiple Unmanned Combat Aerial Vehicles(UCAVs) is one of the key technologies to realize rapid response and effective coordination in the dynamically changing battlefield environment. This dissertation starts from the description and modeling, takes air-to-ground attack and Suppression of Enemy Air Defense(SEAD) mission as the background respectively, under the finite set of distributed control architecture, and focus on the processing of threat area constraints, cooperative constraints and timing sequential constraints of the online collaborative task allocation for multiple UCAVs.Firstly, describe distinguishing feature of the online collaborative task allocation for multiple UCAVs and analyze the control structure of the multi UCAVs cooperative combat system. On these bases, many related concepts and appropriate mathematical model are defined and described.Secondly, taking into account the impact of the battlefield threats, the proposed online route pre-planning method based on improved Rapidly-exploring Random Tree(RRT). A heuristic evaluation function was added to the RRT algorithm, an improved sampling strategy to generate new nodes. A linear regression method is used to solve local optimization problem. Overcome the shortcomings of basic RRT that can't plan the same path repeatedly, we realize an effective compromise between optimal and efficient.Thirdly, focus on the cooperative constraints processing, with a background that multi-UCAV perform air-to-ground attack missions. To tackle the problem that Consensus-Based Bundle Algorithm(CBBA) requires a task to be only assigned to one Agent, reproducing the task that needs multiple UCAVs to attack cooperatively, and then, based on the CBBA framework, it achieves the purpose of a task to be assigned to multiple UCAVs. While the scoring function considering the distance and threat price is designed, online route prediction method based on improved RRT is validated based on the framework of improved CBBA.Finally, in terms of the timing sequential constraints processing, an improvedRecruitment-Based Task Assignment(RBTA)algorithm is designed due to the demand that multi-UACVs are needed to perform SEAD mission cooperatively. On one hand, setup tasks path for each UCAV and allocate task in accordance with the t priority constraints among the tasks sequentially. On the other hand, calculate start time and finish time for each task in the task allocation process, and the UCAV arrival time to the task should not be less than the completion time of its precursor task. It reduced online allocation time for SEAD mission, and the waiting time of multi-UCAVs to perform the same task cooperatively.