機械手臂屬於多輸入多輸出非線性系統,其精確的數學模式難以建立,以模式為基礎的控制器來控制機械手臂是不易實現,因此本研究發展出兩種不需知道受控本體數學模式的智慧型控制器:(1)自組織模糊控制器(2)自組織模糊邏輯與徑向基底類神經網路控制器,以控制二軸與三軸機械手臂系統,分別評估其系統的控制性能。
Robots are multiple-input multiple-output (MIMO) nonlinear systems.It is difficult to identify precise mathematical models for these,making a model-based controller for evaluation impractical. Therefore,this study has developed two model-free intelligent controllers: (1) a self-organizing fuzzy controller,and (2) a self-organizing fuzzy logic and radial basis function neural-network controller.These would be used to control individually a 2-link and a 3-link robotic manipulator to determine control performance.Both intelligent controllers have good control performances in trajectory tracking for robotic motion control,as shown in simulation results.Moreover, the state-space approach was employed in evaluating the stability and robustness of the proposed intelligent controllers. Simulation results have also confirmed that these intelligent controllers have excellent stability and robustness.