人機互動的複合阻抗及導納控制

Abstract

使機器人能在人類生活中更廣泛的使用是一直以來大家研究機器人的目標，然而阻抗控制在剛性環境有較好的表現，而導納控制在軟性環境有較好的表現。因此本篇論文著重在這兩種控制器的整合，並將其整合為切換控制系統，透過模糊任務導向控制及模糊滑模控制來構建切換系統使其適當的在兩控制器間切換。此外，我們整合雙重模糊架構、導納補償器、自適應參數估計到整體架構中以增進整體系統在力及位置的控制。另外，我們使用多個模糊控制來幫助系統根據環境來改變系統參數使機器人能在力及位置上有更好的控制及表現。為了印證我們控制器的可行性，我們使用 Adams-Matlab 偕同仿真來比較我們所提出的控制器和其他控制器的差別，在實驗中我們使用兩種不同的機器人來應用我們的控制器，透過寫字及人機互動來展示我們的控制器使機器人能在剛性及軟性環境皆能有好的表現。

Making the robot widely used is one of the most important topics in robotics research. However, impedance control works well in stiff environments, while admittance control works well in soft environments. Therefore, this thesis primarily focuses on integrating the two controllers into a switch system. This switch system utilizes fuzzy task-oriented control and fuzzy sliding mode control to enable moderate switching between the two systems. Additionally, we integrate a dual fuzzy structure, admittance compensator, and adaptive parameter estimation into the system to enhance the position and force tracking performance of the overall structure. Furthermore, we employ numerous fuzzy control techniques to adjust the parameters based on the robot's environment, thereby improving the robot's performance in position and force tracking. To verify the feasibility of the controller, we conduct simulations using the Adams-Matlab co-simulation to compare our proposed controller with other original control systems. In the experiments, we employ two types of manipulators to implement our controller. We test the manipulators in both soft and stiff environments. The robot writes words and interacts with humans to demonstrate the stiffness and compliance of the robot manipulator.