【计算机科学】【2006.12】基于图的移动机器人路径规划

本文为美国乔治亚理工学院（作者：David T. Wooden）的博士论文，共114页。

本文主要研究现实世界中移动机器人系统的导航、规划和控制问题。第二章包含了本文的第一个贡献，这是对规范的两层混合架构的一种改进：上层考虑规划，下层考虑反应行为。深思熟虑被用来描述更高层次的推理，包括经验记忆和区域或全球目标。反应行为描述了在空间和时间上直接操作机器人信息的低级控制器。在传统的体系结构中，信息是自上而下传递的，协商层向反应层发号施令。第二章介绍了我们在反方向引入反馈的工作，允许行为向规划模块提供决策信息。第三章首先讨论了路径规划问题，特别是由可视图解决的路径规划问题。我们所谓的面向可视图是一种组合规划器，它强调在未知环境中以保证最优为代价进行动态重新规划。给出了单一来源规划的一个例子——目标位置已知且是静态的——将此方法与相关方法（如缩减可视图）进行比较。第四章进一步发展了第三章的工作，将面向可视图扩展到了面向层次的可视图。这项工作直接解决了有向可见度图的一些局限性，特别是在障碍物非凸和障碍物凸壳重叠的情况下失去了最优性。这就产生了一种介于设计用于快速处理动态更新的定向可视图和保证最优路径规划的旧标准降低可视图之间的中间方法。第五章从更高的抽象层次研究路径规划。给定的是一个加权着色图，其中顶点被赋予一种颜色（或换句话说是类），表示与该顶点相关联的环境特征或质量。然后提出问题，“通过这个加权着色图的全局最优路径是什么？”我们用从类和边权重到实数的映射来回答这个问题，并使用Dijkstra算法来计算最佳路径，验证了正确性并强调了具体实现。

In this thesis, questions of navigation,planning and control of real-world mobile robotic systems are addressed.Chapter II contains the first contribution in this thesis, which is amodification of the canonical two-layer hybrid architecture: deliberativeplanning on top, with reactive behaviors underneath. Deliberative is used todescribe higher-level reasoning that includes experiential memory and regionalor global objectives. Alternatively, reactive describes low-level controllers thatoperate on information spatially and temporally immediate to the robot. In thetraditional architecture, information is passed top down, with the deliberativelayer dictating to the reactive layer. Chapter II presents our work onintroducing feedback in the opposite direction, allowing the behaviors toprovide information to the planning module(s). The path planning problem,particularly as it as solved by the visibility graph, is addressed first inChapter III. Our so-called oriented visibility graph is a combinatorial plannerwith emphasis on dynamic re-planning in unknown environments at the expensiveof guaranteed optimality at all times. An example of single source planning –where the goal location is known and static – this approach is compared to relatedapproaches (e.g. the reduced visibility graph). The fourth chapter furtherdevelops the work presented in the Chapter III; the oriented visibility graphis extended to the hierarchical oriented visibility graph. This work directlyaddresses some of the limitations of the oriented visibility graph,particularly the loss of optimality in the case where obstacles are non-convexand where the convex hulls of obstacles overlap. This results in an approachthat is a kind of middle-ground between the oriented visibility graph which wasdesigned to handle dynamic updates very fast, and the reduced visibility graph,an old standard in path planning that guarantees optimality. Chapter Vinvestigates path planning at a higher level of abstraction. Given is a weightedcolored graph where vertices are assigned a color (or in other words class)that indicates a feature or quality of the environment associated with thatvertex. The question is then asked, “what is the globally optimal path throughthis weighted colored graph?” We answer this question with a mapping fromclasses and edge weights to a real number, and use Dijkstra’s Algorithm tocompute the best path. Correctness is proven and an implementation ishighlighted.

1. 研究背景
2. 同时控制欲映射
3. 有向可视图
4. 面向层次的可视图
5. 加权着色图上的全局最优路径规划
6. 结论

更多精彩文章请关注公众号：