本文的研究目標為使用模糊邏輯控制系統,控制輪型行動機器人的運動,結合立體視覺伺服系統,使機器人具有自行搭乘電梯之能力。於行動機器人上安裝四軸機械手臂,分別於機械手臂與行動機器人上架設立體視覺伺服之攝影機。本文將兩組立體視覺伺服系統分工,安裝於機械手臂上之立體視覺伺服系統負責回授控制機械手臂,另一組立體視覺伺服系統則回授控制機器人之運動。 本文使用單一輸入規則模組(single input rule modules, SIRMs)搭配運動重要程度(kinematic importance degrees, KIDs)架構之模糊控制器,降低建立模糊規則庫之困難度,成功設計模糊控制系統,控制機器人到達目標方位;於立體視覺伺服系統,可成功辨識電梯與電梯按鈕,並使用立體視覺計算得到電梯與電梯按鈕於空間中之座標。 於機器人搭乘電梯實驗,結合模糊控制系統與立體視覺伺服系統,使機器人逐步完成自動行走到電梯前、按電梯按鈕、進入電梯、按樓層按鈕與走出電梯之流程。實驗結果,機器人能夠在無人控制的條件下,完成自動搭乘電梯之任務,其完成任務所需的時間約為四分鐘,而完成任務的成功率約為兩成。
The subject of this dissertation is the control of a wheeled mobile robot taking the elevator autonomous based on stereo visual servoing using fuzzy logic. There are two stereo vision systems on the mobile robot. One is the head vision system of the robot used to sense environment for detecting the position of the elevator. The other is mounted on a four-axis manipulator used to recognize the up/down and floor buttons of the elevator. The motion path planning and motion control of the mobile robot are accomplished using fuzzy logic. The processing stage of the mobile robot taking the elevator can be divided into five stages. First, the mobile robot moves to adequate position in front of the elevator door. Secondly, the manipulator recognizes the up/down button and presses it. Thirdly, the mobile robot recognizes the elevator door opened and enters the elevator. Fourthly, the manipulator recognizes the floor button assigned and presses it. Finally, when the elevator reached the target floor the mobile robot recognizes the elevator door opened and leaves the elevator. Experimental results present that the combination of fuzzy logic and stereo visual servoing for controlling the mobile robot can fulfill the task of taking elevator autonomous.