使用光流影像感測的全方向運動平台之空間軌跡追蹤系統設計

Abstract

本論文針對室內環境的輪式機器人提供一套高整合且低成本的定位及控制系統；它可以在一般的室內環境中利用光流影像定位出機器人在空間中的相對方向和位置；同時利用可即時速度控制之空間性軌跡追蹤法來控制該輪式機器人，使得該機器人擁有在多種不同且未知情況的地面上進行速度控制、軌跡追蹤、碰撞偵測以及動態軌跡規劃的能力。本論文亦提出以全向輪為基礎的輪式機器人平台，它於一般的室內環境中擁有較高的運動靈活度；同時我們也整合了嵌入式網路技術，以達成分散式運算及遠端操控機器人之目的。 目前，已經可以透過無線網路控制該平台在多種地形，包含：木板、地毯、磨石地磚以及傾斜的木板之上進行速度控制、軌跡追蹤、碰撞偵測以及動態路徑規劃的任務。 本論文主要分成三大部分，第一部分介紹極值搜尋演算法、軌跡追蹤演算法、光流影像定位演算法以及全方向運動平台演算法。第二部分介紹實驗平台的軟硬體架構。第三部分則是演算法在實驗平台上的實作，結果與分析。

This thesis proposes an integrated and low-cost system of localization and control for an indoor omni-directional wheeled mobile robot using optical flow sensors. The system is capable of commanding the robot to perform velocity control, path tracking, collision detection and dynamic path planning tasks in various unknown landforms. Moreover, the omni-directional wheels offer a higher mobility in an indoor environment, and the embedded Ethernet technology is used to achieve the objectives of distributed computation and remote control. Presently, we can command this platform via wireless network to perform velocity control, path tracking, collision detection and dynamical path planning tasks in all following landforms: plank, carpet, terrazzo floor and inclined plank. The thesis is divided into three parts: The first part is the introduction to the extremum seeking algorithm, the path tracking method, the optical flow localization method and the omni-directional motion algorithm. The second part introduces the overall hardware and software architecture of the platform. The third part is the implementation of overall system on the platform and the discussions and analyses on the experimental results.