Skip to Main Content
In this paper, friction compensation in robotic manipulators is studied. An observer based model of the friction force is utilized for the friction compensation algorithm. In order to evaluate the efficiency of this method two different manipulators with different friction characteristics are examined. A 2-DOF manipulator used for high-speed, micro-meter precision manipulation and a 4-DOF macro manipulator used for long reach positioning task are examined. These manipulators are characterized, according to compensation task classifications (Armstrong, B. et al., 1994), as high-reversal-velocity and low-reversal-velocity tracking tasks, respectively. In each case a steady-state model of friction is experimentally obtained. This model is further utilized in dynamic equations of the friction force during manipulation. It is shown that despite the different nature of the two manipulators the same method can effectively improve the speed and performance of the manipulation in both cases.