Adaptive Control For Robots With Uncertain Kinematic And Dynamic Parameters

In the applications of robot systems, one of the most important factors that impact the performance of the controller is the lack of the model's accuracy. The existing adaptive control methods can effectively solve the control problem of robots with dynamic uncertainties. However, it is quite common that the kinematic parameters are also inaccurate or unknown. Actually the uncertainties of the kinematic parameters often have a greater impact on the accuracy of the robot in cartesian space. Although some studies have attempted to estimate kinematic models by means of computer vision, so far, studies that take both dynamic and kinematic uncertainties into consideration are still rare. In this paper, an adaptive control algorithm for robots with uncertain kinematic and dynamic is proposed. The global asymptotically stable trajectory tracking of Cartesian space is obtained without measuring the velocity of the robot end-effector. Compared with the existing methods, it is with less computation and easier to implement. The simulation case with a two-DOFs manipulator verifies the effectiveness of the method.