| Summary: | 碩士 === 國立中興大學 === 機械工程學系所 === 105 === In recent years, the industrial robot manipulator has been changed wildly used in manufacture. In practical application, it is hard to know the exactly parameters (or models) of the robot manipulator and stiffness coefficient of exterior processing environment since its changes of parameters, exterior disturbances, and friction. As a result, external disturbances and uncertainty of parameters cause difficulty on designing the controller. When the robot manipulator works during the machining, the force control has to be designed in order to smoothly process and avoid damage.
This thesis considers the problem of controlling a redundant robot manipulator in the task space. We discuss the relation between the task space and joint space, and also introduce the knowledge of null space to derive the dynamic model of task and null space. Considering the model system of redundant robot manipulator that is a highly nonlinear dynamic model, with the interaction interference and uncertainty, we utilize the position and force controller design to overcome these problems by fuzzy neural network. Meanwhile, the stiffness coefficient of the environment is unknown. Therefore, we use the gradient descent method to estimate the stiffness coefficient of environment to achieve the adaptive force control. The stability analysis of the closed-loop system and the corresponding update laws are given by Lyapunov stability theorem. Finally, we employ our proposed control scheme in the redundant robot manipulator KUKA LWR4 with 7 –DOF to illustrate the performance and effectiveness.
|
|---|
