【作者】 聂鑫；

【导师】 王耀南； 余洪山；

【作者基本信息】 湖南大学 ， 控制科学与工程， 2011， 硕士

【摘要】 随着科学技术的发展,人类的研究和活动领域已由陆地扩展到深海和太空。利用移动机器人进行空间探测和开发,已成为21世纪世界各科技发达国家开发空间资源的主要手段之一。移动机器人在月球和火星等外星球表面导航时,将面临复杂的未知环境。尽管在移动机器人导航控制理论和方法的研究中,确定性环境的导航控制方法已取得了大量的研究和应用成果,但是却不能直接应用于机器人处于未知环境中的自主导航。因此本文针对未知环境中的移动机器人的自主导航的问题展开研究,提出了基于激光雷达的自主导航系统,经过实验以及仿真研究证明了该系统具有一定的前瞻性和可应用性。首先,本文概述了移动机器人的定义与发展现状、主要应用领域和典型代表机器人。综述了移动机器人自主导航技术的研究现状,分析移动机器人自主导航中所面临的难点问题,介绍了现有的一些自主导航方法并分析了其利弊。其次,本文介绍了基于激光的移动机器人自主导航系统的系统设计,首先详细介绍了自主导航系统的硬件系统的各个组成部分,包括了Pioneer2移动机器人平台、激光测距仪、惯性测量系统。其次介绍了所开发的软件功能模块组成、基本功能和工作流程。然后,本文介绍了基于激光测距仪的室内环境建模。首先针对研究的室内的环境特点,对激光传感器进行了数学建模。接着介绍了直线检测与拟合的方法。引入了Hough变换对激光扫描的距离数据进行分簇,并用最小二乘法进一步拟合线段,然后构建出室内环境线段特征地图的方法。接着介绍了软件的界面设计,代码设计,相关实验等内容。最后,主要介绍了移动机器人在自主导航中的路径规划方面的研究,讨论了现有的各种现有路径规划的方法的优缺点,分析了至今还未很好解决的问题。针对一些复杂障碍物提出了一种基于行为的路径规划方法,将整个路径规划行为分为逼近目标行为、避障行为、子目标规划行为。通过合理的行为转换规则控制机器人在几种行为中切换与相互配合来完成路径规划任务。机器人在同一时间内只采用其中一种行为。接着分别详细介绍了几种子行为的方法,并在其后进行了仿真实验,证明了该方法的有效性。接着介绍了两种移动机器人的慎思规划方法作为后续的研究方法,对其适用领域与实现方法进行了详细介绍。更多还原

【Abstract】 With the development of science and technology, the fields of human studies and areas of human activity have been extended to the deep sea and land space. Using mobile robots for space exploration and development has become one important way to develop space resources in worldwide technology developed countries of 21st century. When navigate on the surface of alien planet, such as moon and Mars, mobile robots will face a complex unknown environment. Although in the study of mobile robot navigation control theory and methods, a lot of research and application results of navigation and control approach under deterministic environment had been made, but they can not be directly applied to the mobile robot navigation in unknown environments. So this paper mainly researched the problems of the robot’s autonomous navigation under complex unknown environments. This paper proposed a kind of mobile robot navigation system based on laser radar and proved that this system had certain forward-looking and can be applied.First, this paper outlined the definition, the development status, the main applications and a typical representative of mobile robots. And, this paper summarized the research status of autonomous navigation for mobile robot, analyzed the problems which was faced in autonomous navigation for mobile robot,and introduced same existed means of autonomous navigation and their pros and cons .Next, the design of the laser-based mobile robot autonomous navigation system was presented in this paper. Firstly, the various components of autonomous navigation hardware system, including the Pioneer2 mobile robot, laser range finders, inertial measurement systems, were detailed. Then, the composition of the designed software function module, basic function and software workflow were presented.Then, the indoor environment modeling based on the laser range finder was introduced. Firstly, laser’s Mathematical model was made for the features of the indoor environment. Secondly, the line detection and fitting methods were introduced. A indoor environment feature mapping way which clustered the range data from the laser and fitted lines with least squares was introduced. Then, this paper detailed the interface design, program design of the navigation software system and some related experiments. Finally, this paper introduced some researches on the path planning of the mobile robot autonomous navigation., discussed the pros and cons of the various existed path planning methods, and analyzed some problems unresolved until now. Aiming at some complex unknown obstacle, this paper proposed a path planning method based on behavior control which separated the entire path planning behavior into 3 sub behaviors which was approximation behavior, Obstacle Avoidance behavior, and sub goal planning behavior. Rational behavior change rules were set to control robot change behavior among the 3 rules and cooperate each behavior to finish the navigation task. Mobile robot can only use one behavior at one time. Then this paper detailed the 3 sub behaviors, and did some simulation which proved that this algorithm is effective and verified the applicability of the proposed method for some complex environments. Then 2 deliberation path planning methods for mobile robot were introduced for further research, and their application field and realization method were detailed.更多还原