Static and Dynamic Friction Models for Robotic Manipulators: State of the Art and Experimental Comparison

Achieving an accurate and efficient dynamic modelling is fundamental in many robotic applications. The effect of friction in the dynamic behaviour of a robot is one of the most challenging factors to predict, due to the complexity of the several phenomena that cause it. Several friction models are available in the literature, each of which can be more or less suitable depending on the situation and working conditions. In this paper, we present the state of the art and experimental comparison of friction models for robotic manipulators. We first consider both static and dynamic friction formulations to evaluate their ability to consider the different friction effects and the factors that influence friction behaviour. Then, twenty-five selected models are experimentally compared to evaluate their ability to predict friction effects in a UR5e robot with six degrees of freedom. The results of extensive experiments show good performance of most of the considered friction models in predicting the friction torque, and in evaluating the energy consumption of the robot when applied together a complete dynamic formulation. The findings of this study can be used as guidelines for determining the best friction model for the operating conditions under consideration.