本論文使用螞蟻演算法(Ant Algorithm)和模糊控制(Fuzzy Control)，設計一台可以在辦公室環境中自由行走的輪型機器人，並透過基因演算法來調整相關參數使其達到較佳的狀態。 論文中先利用辦公室環境地圖，搭配螞蟻演算法找出出發點與目標點間最佳路徑，接者利用模糊理論去控制輪型機器人，使其依循此路線行走，當遇到障礙物時會避開障礙物接者繼續追蹤路線行走，如此一來就能依循路線直到抵達目標點，文中並設計一個方法來避免輪型機器人陷入死路無法達成任務的情況。當抵達目標點後，並依照此路線上遇到的障礙物，根據障礙物特性給予不同的懲罰值，使得之後出任務時在使用螞蟻演算法進行路徑規畫時，能夠更有智慧的尋找出最佳路徑。 在實作的部分，本論文利用雷射測距儀獲得障礙物與輪型機器人間的相對位置，利用位置控制器記錄馬達旋轉的資料，並在C++ Builder 上撰寫路徑規畫、避障及追蹤控制器等程式。最後透過定位控制，路徑規畫與追蹤和閃避障礙物等實驗來驗證本論文中各種方法之有效性。 This paper uses a kind of ant algorithm and the fuzzy control to design a wheeled mobile robot(WMR) that can move around in the office environment, and through the Genetic Algorithm to adjust the parameters to achieve a better state. First, use the ant algorithm to find the best path between the starting point and the target point by office map. And then design a WMR via the fuzzy control. The purpose is to make the WMR have perform the path tracking and to avoid obstacles. The study also designs a method to avoid the WMR into the dead end and can’t achieve the task. After completion of the task, it will adjust parameters based on characteristics of obstacles. This method makes the WMR have more intelligent in the path planning. In the experiment, use a laser range finder to apply environmental data from the wheeled mobile robot, and then use a position controller to find the information for the motor. Finally, simulations and experiments are given to verify the feasibility and effectiveness.
