本論文針對受約束機械臂控制系統提出一種適應性T-S模糊輸出回授追蹤控制方法。對於系統具參數不確定或結構未知時,設計適應性T-S模糊觀察器估測其系統狀態,再應用觀察器遞回的方法設計控制器。另外考慮致動器動態響應之機械臂,亦可利用相似的T-S模糊觀察器遞回控制器設計達到運動和力量之追蹤控制。利用Lyapunov穩定性分析證明觀察器和控制器皆能為一致最終有界uniformly ultimately bounded (UUB)的情況。最後,利用MATLAB Simulink和Real Time Workshop以及dspace DS1104控制卡實現一個實際機械臂控制實驗,並藉此驗證本論文所設計之控制方法的可行性。從理論、模擬結果和實作成果驗證本論文的設計方法具有良好追蹤能力和快速收斂的成果。
In this thesis, an adaptive T-S fuzzy output feedback tracking control is proposed for constrained robot manipulators. To deal with unknown parameters or system uncertainties, an adaptive T-S fuzzy observer is designed for state estimation, then the controller is designed based on observer backstepping. Moreover, for the consideration of actuator dynamic response for robot manipulators, we can achieve motion and force tracking control using the similar T-S fuzzy observer backstepping controller design. From Lyapunov stability analysis, we prove that the observer and controller satisfy the uniformly ultimately bounded stability. Finally, an experiment of actual robot manipulator control is carried out by Matlab Simulink, Real Time Workshop, and dspace DS1104 control card to verify the feasibility of the proposed control methods. From both simulations and experiments, we can conclude that the design of this thesis has a good tracking performance and fast convergence.