本論文使用階層式模糊邏輯控制器（Hierarchical Fuzzy Logic Controller，HFLC）設計輪型行動機器人（Wheeled Mobile Robot，WMR）的控制器，並配合模擬退火法（Simulated Annealing，SA）將此控制器之相關參數調整到一個較佳的狀態，來對WMR作運動控制，使WMR 能從任意初始位置及方位角完成定位控制與軌道追蹤控制。 另外本論文也提出兩種不同狀況之避障路徑系統，第一種為當環境中的障礙物為特定形狀時，利用模糊理論即時決定一個能閃避障礙物的最佳轉向角度。第二種為當環境中的障礙物為任意形狀時，便結合子目標選擇法及曲線插補法設計一路徑規劃法，求得一個能避免與障礙物碰撞之最佳路徑，使WMR不但可以順利的避過障礙物，且能以最短路徑到達目標位置。 在實作驗證部分，本論文利用影像處理技術擷取輪型行動機器人與障礙物之狀態，並以Borland C++ Builder為平台撰寫控制器及避障路徑規劃等演算法，也利用單晶片Microchip PIC16F877進行馬達轉速控制，而整個輪型機器人之操控及狀況判斷與處理均建構在視窗環境下。最後透過定位控制、軌道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 thesis uses the hierarchical logic fuzzy controller (HFLC) to design controllers for a wheeled mobile robot (WMR). Partial parameters of the proposed controllers are fine-tuned by simulated annealing (SA) for better performance. The proposed controllers are applied to the motion control of the WMR from any arbitrary position and direction to achieve the position control and the trajectory tracking control. There are also two path-planning methods for obstacle avoidance has been proposed in this thesis. The first one is based on the fuzzy theory to decide an optimal steering angle for collision avoidance of the WMR real-time if the shapes of the obstacles in the environment are all particular. When the shapes of the obstacles in the environment are arbitrary, we combine the subgoal selection method and the spline interpolation method to design a path-planning method to obtain an optimal path for obstacle avoidance and reach the target position. In the part of the implementation, we use the technology of the image processing to acquire the movements of the WMR. In order to operate the WMR, the control algorithm and the path-planning method are designed with Borland C++ Builder and the motor speed of the two wheels is controlled by microchip PIC16F877. All of the operations of the WMR are established in the circumstance of the windows. Finally, the experimental results are given to demonstrate the effectiveness of the proposed methods in this thesis.
