本論文提出一個新穎的T-S模糊系統之模糊控制器設計方法，此方法之設計概念是根據基因演算法與線性矩陣不等式理論基礎架構而成，藉此設計方式，模糊控制器的迴授增益可不經由繁複的數學推導及計算而自動地被尋得，論文中也將此設計方法應用在輪型行動機器人的軌跡控制問題上。 在本論文中，首先對使用線性矩陣不等式方法求解的輪型行動機器人軌跡控制模糊控制器所造成不良影響做一敘述，並介紹如何以本論文提出之設計方法來改善，且在論文中以模擬及實作的方式呈現其控制效果，確認本論文所提出之結合基因演算法與線性不等式設計方式的有效性及正確性。 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動態，並在C++ Builder撰寫模糊控制器程式及利用單晶片PIC16F877進行馬達轉速控制，而整個輪型機器人之操控及訊號處理均建構在視窗環境下。 A novel design method for T-S fuzzy controller is proposed in this thesis. The method is based on the genetic algorithm and linear matrix inequality. The controller feedback gains can be found automatically without complex mathematical calculation via this approach. The approach is applied to the trajectory control of a Wheeled Mobile Robot. In this thesis, we describe the ill influences of trajectory control of WMR’ fuzzy controller by using LMI method. We also introduce how to improve the effect by the proposed method. The results are shown by simulations and experiments with WMR to ensure the feasibility and the validity of our proposed method. In the part of implementation, we use the technology of image process to acquire movements of WMR. In order to operate WMR, the fuzzy control program was designed by C++ Builder and motor speed is controlled by microchip PIC16F877. All of the operations of WMR and signal processes are designed under the circumstances of windows.
