CASIA OpenIR  > 毕业生  > 硕士学位论文
面向SEAD任务的多无人机协同任务规划研究
其他题名Collaborative Mission Planning of Multi-UAV based on Military Suppression of Enemy Air Defenses(SEAD)
陈艳红
学位类型工学硕士
导师常红星
2014-05-26
学位授予单位中国科学院大学
学位授予地点中国科学院自动化研究所
学位专业计算机应用技术
关键词无人机 防空压制 协同任务分配 实时航路规划 Unmanned Aerial Vehicles Suppression Of Enemy Air Defenses Collaborative Mission Planning Real-time Route Planning
摘要多无人机协同任务规划能够充分发挥多无人机的协同作战优势,不仅能够使多类型任务与多性能无人机之间保持良好的协调性,同时当遭遇突发威胁或出现无人机战损时,能够快速完成任务再规划,使任务执行效率最大化。多无人机协同任务规划是一个复杂的多约束、多目标优化问题,通过优化算法为多无人机解算出协同的任务计划,使多无人机协同作战的整体效能优于各个无人机单独作战效能的总和。本文以多类型任务、多无人机作战平台的防空压制任务(Suppression of Enemy Air Defenses,SEAD)为背景,对多无人机协同任务规划问题的两个关键问题——复杂约束下多无人机协同任务分配和动态环境下无人机实时自主航路规划问题展开研究。论文主要工作包括: 1.研究多无人机协同任务规划的系统结构,分析了SEAD任务中多无人机平台特点及其同SEAD任务之间的耦合关系、影响多无人机协同任务规划的约束条件,描述了SEAD任务特性(时间窗、执行者约束等)以及任务的优化指标,进而建立了多无人机协同任务规划的数学模型。 2.提出了基于分布式并行协商机制的算法求解多无人机协同任务分配问题。针对多无人机协同任务分配这一复杂的多约束组合优化问题,采用多Agent并行协商的分布式任务规划方法,各无人机根据自身性能和任务属性对待分配任务集合进行自主选择和安排,多无人机之间以较少的燃油消耗、合适的任务执行时间窗及最大化任务收益协商任务分配结果,从而将不同类型的任务以合适的任务执行时间窗分配给合适的无人机。实验结果表明该方法对于多约束条件下的任务分配问题具有快速收敛性,且计算复杂度低。 3.提出了一种基于行为决策的动态环境下无人机在线自主航路规划方法,将无人机实时航路规划分解为障碍物规避和向目标点前进两种行为方式,并提出了避障导引速度和向目标点前进导引速度计算方法。针对无人机作战环境复杂多变、情报信息准确性和精确性欠佳的情况,利用无人机实时探测到的局部环境信息及全局目标位置信息,在每一时刻的规划窗口内,通过导引速度计算确定无人机的期望飞行方向,根据无人机当前飞行方向与期望方向的差异实时做出决策,导引无人机避开障碍物,同时朝向目标点飞行。实验结果表明该方法可以快速规划航路点,并具有灵活避障能力。
其他摘要Teams of Unmanned Aerial Vehicles (UAVs) take good advantage of the superiority of co-operative operations. Collaborative mission planning of multi-UAV can not only make the UAVs’ capabilities match the multi-mission’s needs, but also re-plan quickly when encounter with sudden threats or loss of aircrafts. Collaborative mission planning of multi-UAV is a complex optimization problem with multiple constraints and optimization objectives, it figures out the cooperative operation plans, which aims at maximizing the whole UAVs’ combat effectiveness. This thesis focuses on collaborative mission planning for multiple autonomous UAVs performing Suppression of Enemy Air Defenses (SEAD) missions. The research work contains the two aspects: collaborative tasks allocation and real-time route planning. The main contributions are as follows: (1)Study the system construction of Collaborative Mission Planning problem. Analyze the main constraints and optimization objectives. Define the main types of missions and UAV platforms by studying representative scene of SEAD. Describe the key properties of task allocation and route planning, and finally formulate the mathematical model of Collaborative Mission Planning problem. (2)A decentralized planning method based on negotiation-mechanism is proposed to solve the multi-task’s allocation, which is a complex and combinatorial optimization problem. Each UAV automatically construct its own task sequence based on its capability and flight situations, multi UAVs negotiate for the final allocation of tasks based on the fuel penalty, timing constraints and proceeds of performing the task, therefore assign the tasks to proper UAVs and executed at proper times. The simulation results prove the effectiveness and rapid convergence of the proposed task allocation method. (3)A new real-time route planning algorithm based on autonomous behavior decision is proposed. The process of online route planning can be come down to two behavior modes – Obstacles Avoidance and Towards the Goal. New computing methods of guide velocity are proposed. Based on the fact that the information of the flight environment is partly unknown or even totally unknown, when UAV plans its route in real-time, it makes use of the local environment information which detected online and the global goal information which loaded before taking off(or renew during the flight), the route planning algorithm compute the expected guide velocity according to the re...
馆藏号XWLW2052
其他标识符201128014629072
语种中文
文献类型学位论文
条目标识符http://ir.ia.ac.cn/handle/173211/7724
专题毕业生_硕士学位论文
推荐引用方式
GB/T 7714
陈艳红. 面向SEAD任务的多无人机协同任务规划研究[D]. 中国科学院自动化研究所. 中国科学院大学,2014.
条目包含的文件
文件名称/大小 文献类型 版本类型 开放类型 使用许可
CASIA_20112801462907(2144KB) 暂不开放CC BY-NC-SA请求全文
个性服务
推荐该条目
保存到收藏夹
查看访问统计
导出为Endnote文件
谷歌学术
谷歌学术中相似的文章
[陈艳红]的文章
百度学术
百度学术中相似的文章
[陈艳红]的文章
必应学术
必应学术中相似的文章
[陈艳红]的文章
相关权益政策
暂无数据
收藏/分享
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。