A Smoothed GMS friction model suited for gradient-based friction state estimation

The multi-state Generalized Maxwell-Slip (GMS) friction model is known to describe all essential friction characteristics in presliding and sliding motion. It is also known that due to its switching state conditions between presliding and sliding, gradient-based parameter and state estimation methods cannot be implemented efficiently. Efficient on-line state and parameter estimation is essential for model-based friction compensation in order to track changes of friction characteristics in time and space. This paper presents a Smoothed GMS (S-GMS) friction model with an analytic set of differential equations well suited for gradient-based estimation techniques. Friction state estimation is implemented with an Extended Kalman Filter (EKF) to validate the S-GMS model numerically and experimentally, and compare its behavior with the GMS model.