【作者】 周其忠；

【导师】 闫利；

【作者基本信息】 武汉大学 ， 大地测量学与测量工程， 2010， 博士

【摘要】 在未来作战体系中,往往通过使用多个无人作战飞行器(UCAV)对目标进行协同攻击,以完成特定的作战任务。为满足这种作战需求,如何合理分配我方可使用的UCAV资源,在复杂的飞行环境中为每枚UCAV生成有效的航迹,并且协调各UCAV的到达时间是一个在未来作战应用中急需解决的难题。针对这一难题,本文首先分析了长航时UCAV飞行特点以及飞行环境,研究了稀疏A*算法(SAS)和遗传算法(GA)的相关理论,在此基础上,结合递阶控制理论,提出了一种解决复杂环境下多UCAV协同规划问题的分层递阶体系框架,并就该体系框架的相关内容进行了系统性研究。这些内容包括复杂飞行环境建模、航线规划建模、三维航线规划及多航线规划算法、任务分配与协同技术等。最后建立了相应的原型系统,并进行了算例试验验证。试验结果表明,在该体系框架下,可以有效地对UCAV资源进行分配,给出满足长航时精确导航UCAV复杂约束约束的多航线,UCAV按给定的航线飞行,在相互不碰撞情况下,同时到达目标,实现给定的作战任务。因此,该体系框架能有效解决复杂环境下多UCAV协同规划的难题。归纳起来,论文的主要贡献和创新点如下：1)系统地提出了复杂环境下分层递阶多UCAV协同规划体系框架,将复杂环境下多UCAV协同规划过程划分为航线预规划、任务分配、航线规划与任务协同四个逻辑上互相关联与迭代的过程,逐层降低了问题的复杂度,有效解决了复杂环境下多UCAV协同规划技术难题。2)提出了独立度矩阵的新概念,在此基础上,建立了反映航线之间碰撞程度的时空数学模型。在航线规划过程中,通过基于空间独立度矩阵的碰撞模型对UCAV碰撞情况的定量化表达,完成了多航线的规划；在任务分配或任务协同过程中,在确定发射时刻的同时,结合UCAV的飞行速度及惯导、高度漂移规律,利用基于时空独立度矩阵的碰撞模型,有选择性地进行时间控制,解决了UCAV飞行过程中的避碰问题。3)提出了一种基于全矢量数据三维航线规划方法,该方法可在考虑UCAV机动性能约束的情况下快速给出三维航线。4)提出了基于GA&SAS混合算法的大范围复杂环境下三维多航线规划方法。该方法可充分利用SAS的强启发遍历性搜索特点和GA算法的强空间搜索能力,以及基于矢量数据的三维航线快速算法给出的引导信息,来实现三维多航线规划。5)建立了基于多发射阵地、多UCAV、多目标协同作战的任务分配数学模型,结合遗传优化理论,提出了基于遗传的启发式任务分配算法,该算法可以快速给出满足指挥人员决策意图的任务分配结果。6)在考虑发射阵地连续发射UCAV的时间间隔要求以及UCAV的速度调整能力前提下,提出了基于时间控制的任务协同方法进行UCAV之间的防碰撞处理。该方法算法简单,通用性强,能够有效的解决UCAV之间的防碰撞问题。本文的研究结果为解决复杂环境下多UCAV协同规划问题提供了较好的解决方案。更多还原

【Abstract】 UCAVs (Unmaned Combat Aerial Vehicles) are often used to strike targets coordinately to meet operation requirements. How to allocate UCAVs available, generate their routes, and set time of arrival for them is a difficulty to achieve goals of operation successfully.In order to find answers to it, an architecture oriented to multi-UCAVs’ mission planning in complex combat environments has been proposed in the paper. Much theoretical research work on the aspects related to the architechure,including complex flight environment model,route planning model,3D route planning algorithm and multi-route planning algorithm,has been conducted systematically.A large number of computer simulations has been carried out with the prototype developed for the architecture.The experiments shows that An operation scheme can be attained under the architecture by considering task force availabilities, route planning in complex environments and collision avoidance among UCAVs. These initiatives include:1) A mission planning hierarchical framework which logically integrates route preplanning, task assignment, route planning, and task coordination model. It divides multi-UCAVs’mission planning problem into three logically-related levels-route preplanning,task assignment,route planning and task coordination. This can lessen problem constraint complexities and solve the difficulties of multi-UCAVs’mission planning effectually.2) An independent matrix concept which can explain quantitatively collision between any two routes has been proposed. It has been used in the process of multi-route planning, task assignment and coordination. Expected routes can be chosen to be adjusted, or be formed a route group which meets criteria of collision avoidance.3) A vector model-based 3D route planning algorithm has also been presented. It can rapidly generate 3D routes under UCAV’s maneuvering constraints.4) A GA&SAS mixed 3D multi-route planning method has been proposed. A batch of routes which meets complex constraints of long-range UCAV with precise navigation systems can be generated by taking advantages of SAS’s grid-by-grid search and GA’s global random search abilities.5) A task assignment model has been built by correlating launch sites, UCAVs and targets, and a heuristic GA based task assignment algorithm has been proposed. The algorithm can give expected assignment results rapidly.6) A time-control based task coordination method which makes UCAVs’ collision avoid effectively has also been presented by considering UCAVs’launching time intervals and speed adjustment capabilities. It makes UCAVs’collision avoid, and minimize launch time window under the TOA (Time Of Arrival) constraint.Therefore, the research work presented in the paper provides a solution to multi-UCAVs’coordinated mission planning in complex environment.更多还原