摘要
感知任务的合理分配是影响无人机目标区域覆盖的重要因素,针对任务需求差异并考虑无人机局部观测性和环境不确定性,提出一种面向目标区域协同覆盖的感知任务分配方法。将目标区域进行差异划分,构建基于分布式马尔可夫覆盖模型的任务分配控制框架;利用目标线路集和任务扩散调度序列集对目标区域进行差异化计算,并提出基于强化学习的任务差异化分配方法,实现动态目标区域的最优覆盖策略。仿真实验结果表明:在满足航向速率和空速的条件下,任意两台无人机之间可以合理地扩散调度任务,同时通过差异化学习方法使覆盖线路代价和目标函数适应值收敛稳定且覆盖率达到90%以上,实现对任务分配的有效控制。
The reasonable allocation of tasks is an important factor affecting target area coverage unmanned aerial vehicle(UAV).Focus on the issue of requirement coverage,we propose a method of task allocation control of UAVs for targets area coverage.It used the division of voronoi diagram to divide target area in variation and built a task allocation control framework based on distribute Markov model.And then it computed variation for target area with roads set and diffusion scheduling sequence set,and proposed an algorithm of reinforcement learning based on variation for task allocation that could realize optimal coverage strategy for dynamic target area.The simulation results show that the task between any two UAVs can be scheduled reasonably in the case of unknown conditions and dynamic targets,while the plant cover line cost and target function adaptive value can be rapidly convergence to the stable value and the coverage rate reaches more than 90%that implements effective control over task allocation.
作者
宋伟中
王行业
王宁
Song Weizhong;Wang Xingye;Wang Ning(School of Information Engineering,Huanghe S&T University,Zhengzhou 450000,Henan,China;School of Information Engineering,North China University of Water Resources and Electric Power,Zhengzhou 450000,Henan,China)
出处
《计算机应用与软件》
北大核心
2021年第5期75-81,共7页
Computer Applications and Software
基金
河南省科技攻关计划项目(172102310634)。
关键词
多无人机
目标区域覆盖
任务差异化分配
扩散调度
Unmanned Air Vehicle(UAV)
Target area cover
Task differentiation allocation
Diffusion scheduling
