【作者】 赵宏伟；

【导师】 周航； 杨国庆；

【作者基本信息】 南京航空航天大学 ， 交通运输规划与管理， 2021， 硕士

【摘要】 针对末端物流出现的配送不及时和人力成本高等问题,学者提出了使用旋翼无人机完成末端物流配送。目前,学者对无人机路径规划的研究多在准三维空间,以距离或时间最短为目标函数进行规划,在三维空间路径规划的时空复杂度较二维呈指数增长,该类方法无法满足快速规划要求,同时,由于没有考虑自然环境的影响,导致规划的路径不能满足实际配送需求。针对以上问题,本文研究了自然风影响下旋翼无人机的力学模型并构建了悬停和巡航两种状态的能耗最低模型以保证配送路径满足能耗要求。研究了太阳运动规律并设计了光照代价最低模型以减少光照对无人机视觉导航的影响。为了进一步提高路径规划的效率,本文在建立包括悬停精度和能耗等约束条件基础上,设计了改进蚁群-A~*混合算法完成旋翼无人机物流配送实时路径规划。仿真结果表明,在3m/s西风影响下的环境中,采用距离最短模型规划的路径实际能耗消耗超过了电池最大能耗,不能完成指定配送任务。而以能耗最低模型规划的路径比距离最短模型规划的路径能耗减少了10210.311 J(1.8%),并能顺利完成配送任务。以光照代价最低模型规划的配送路径比距离最短模型规划的路径光照代价降低了480.3 rad·s(41.94%),对无人机基于视觉的远程控制有更好的视觉效果。改进后的A*算法比传统A~*算法运行时间减少了83.5%,改进后的蚁群-A~*路径规划算法比A~*-蚁群算法减少了95.6%。本文提出的路径规划模型有效解决了物流配送路径可行性的问题,改进的混合算法解决了物流配送路径规划的实时性问题。更多还原

【Abstract】 Because of the problems of untimely distribution and high labor cost in the terminal logistics,scholars proposed to use rotor UAV to complete the terminal logistics distribution.At present,scholars’ research on UAV path planning is mostly in quasi-three-dimensional space,taking the shortest length or time as the objective function,and the space-time complexity of path planning in three-dimensional space is exponentially higher than that in two-dimensional space.This kind of method can not meet the requirements of rapid planning.At the same time,due to the lack of consideration of the impact of the natural environment,the planned path can not meet the actual distribution needs.Because of the above problems,this paper studies the mechanical model of rotor UAV under the influence of natural wind and constructs the lowest energy consumption model of hovering and cruising to ensure that the distribution path meets the energy consumption requirements.The law of solar motion is studied and a model with the lowest illumination cost is designed to reduce the influence of illumination on UAV visual navigation.To further improve the efficiency of path planning,based on the establishment of constraints including hovering accuracy and energy consumption,an improved ant colony-A* hybrid algorithm is designed to complete the real-time path planning of logistics distribution of rotor UAV.The simulation results show that in the environment under the influence of 3 m/ s westerly winds,the actual energy consumption of the path planned by the shortest length model exceeds the maximum energy consumption of the battery and can not complete the specified distribution task.Compared with the shortest length model,the path energy consumption planned by the lowest energy consumption model is reduced by 10210.311 J(1.8%),and the distribution task can be completed successfully.Compared with the shortest distance model,the distribution path planned by the lowest lighting cost model reduces the lighting cost by 480.3 rad ·s(41.94%),and has a better visual effect on the visionbased remote control of UAV.The running time of the improved A~* algorithm is 83.5% less than that of the traditional A~* algorithm,and the improved ant colony-A~* path planning algorithm is 95.6% less than that of the A~*-ant colony algorithm.The path planning model proposed in this paper effectively solves the problem of the feasibility of logistics distribution path,and the improved hybrid algorithm solves the real-time problem of logistics distribution path planning.更多还原