Static and dynamic friction compensation in trajectory tracking control of robots

Abstract

This paper deals with friction compensation techniques for accuracy trajectory tracking control of robots. We extend the traditional model-based adaptive control algorithms that are prevalent in the robotics literature with adaptive friction compensation. Adaptive controllers that consider both static and dynamic friction effects in the robot dynamics are proposed. Stability of the proposed adaptive controller with friction compensation is shown. An extensive experimental study has been done to illustrate the importance of friction compensation in trajectory tracking. The tracking performance of the proposed adaptive controller utilizing the dynamic friction models for compensation is far superior to the conventional adaptive controller without friction compensation. Using the data collected for a series of experiments we compare the proposed adaptive controllers with static and dynamic friction compensation with two well known robot control algorithms, i.e., model-based computed-torque control and adaptive control. A representative sample of the experimental data is shown and discussed.