球與平板系統主要是藉由平板角度的變化導致平板上方的球接受到不同方向的重力影響，藉此改變球在平板上的位置，整個控制系統是典型的多變數、不確定性、非線形系統，由此可知整個球與平板系統的動態且非線性特性呈現出令人感興趣的控制研究問題。本論文分別考慮一維的球與桿系統與二維的球與平板系統，其中所設計的二維球與平板系統機構是擁有三自由度的3-RRS結構，透過德州儀器公司所生產的TM4C123GXL LaunchPad開發板進行PD控制演算法、模糊控制演算法與逆向運動學來計算三顆馬達的控制量並藉此改變桿或平板的傾斜角度，本論文還設計了一個使用者人機界面方便操作者使用，其中使用了OpenCV函式庫來做影像處理，透過把RGB色彩模型轉換成HSV色彩模型機構後，再進行灰階與二值化處理，最後在求出球體輪廓並計算球的坐標。最後，經由實際實驗結果可發現，模糊控制演算法在該機構上效果優於PD控制演算法，可以較快的抵達定點且位置較為精準。 Changing the plate’s angle to lead the ball which on the plate receives the gravity from the different directions to change the positions, it’s so called the Ball and Plate System, the typical multivariate, uncertain and non-linear coupling system. It can be seen that the dynamic and non-linear properties of the ball and plate system shows the interesting control research issue. It’s both discussing the one-dimensional ball and pole system and the two-dimensional ball and plate system in this paper, especially focusing on the two-dimensional ball and plate system which has 3-RRS structure with 3-DOF. It uses TM4C123GXL LaunchPad EVB which manufactured by Texas Instruments Inc. doing the PD control algorithm, the fuzzy control algorithm and the inverse kinematic to calculate the control quantities of the three motors and move the mechanism. And it’s also designed the user HCI in this paper, which uses the OpenCV library to do image processing, through the RGB color model switch into HSV color model, then doing gray scale and binarization. Finally, drawing the sphere contour and calculate the coordinates. This method can help the operator easy to use. Through the experimental results, we can find that the fuzzy control algorithm is superior to the PD control algorithm, and can reach the position faster and more accurate.
