本研究旨在利用模糊控制理論來發展輪型機器人的防撞系統，此系統以Labview圖控軟體作為主控制器，利用Labview內建的的資料擷取(DAQ)及控制元件，透過Wi-Fi對輪型機器人進行資料讀取與命令下達。為將模糊控制理論融入主控制器中，使輪型機器人能與前方物體保持安全的相對距離，我們將讀取到的紅外線測量距離之相對電壓，作為模糊規則前件部的輸入，然後利用模糊推理機制，計算理想的馬達速度。為達成此目標，本研究利用Labview，將輪型機器人回傳的數據，存入資料夾中，再由Matlab讀取此資料夾的電壓值，然後Matlab進行曼達尼乘積推理機制計算，得到理想的馬達速度，來確保輪型機器人與物體之間防撞。為驗證此理論之可行性，此研究以Festo Robtino輪型機器人作為實驗平台，以Labview與Matlab混合撰寫程式，而在距離的模糊化方面，定義三個正規化的模糊歸屬函數作為分類，分別為4~10公分、10~30公分及30公分以上，實驗結果顯示，Festo Robtino輪型機器人具有良好的防撞功能。
A fuzzy-based collision avoidance system is developed for wheeled robot in this study. In the proposed system, Labview is served as the main controller to take advantages of its built-in data acquisition (DAQ) and control elements. The data and command access between wheeled robot and computer is progressed via wireless Wi-Fi. In order to maintain a safe relative distance between the wheeled robot and forward object, a fuzzy controller is embedded into the main controller. The feedback voltage, which is proportional to the infrared ray measuring distance, is treated as the input of the antecedent of fuzzy rules and then the fuzzy inference mechanism is used to calculate the desired motor speed. To achieve this goal, Labview is programmed to read the feedback data and store the data into a folder. Then, Matlab is programmed to read the stored data and the Mandani product reasoning mechanism is designed to calculate the ideal motor speed to ensure the anti-collision between the wheeled robot and the object. Finally, to verify the practicability of the proposed control scheme, the Festo Robtino wheeled robot is used as the experimental platform, while the Labview and Matlab programs are mixed to construct the controller in this study. The measuring distance is fuzzified by three normalized fuzzy membership functions, 4 to 10 cm, 10 to 30 cm and above 30 cm, respectively. From the experimental results, it shows that the Festo Robtino wheeled robot possesses good anti-collision performance.
