Comparative study on friction compensation using Coulomb and Dahl models with extended and unscented Kalman filters

In this paper, the Coulomb and the Dahl friction models have been simulated and estimated using extended Kalman filter (EKF) and unscented Kalman filter (UKF). The estimated friction is used by a controller to compensate and track a desired sinusoidal position profile. In the simulation, the real friction is estimated using EKF and UKF with the Coulomb and the Dahl friction models. The designed controller consists of both feedback and feedforward for position tracking. The proportional and derivative (PD) output feedback is selected. The desired trajectory inertia and the estimated friction are compensated as the feedforward. The tracking performance is analyzed in terms of the normalized root mean square error. It is found that the UKF based controller using Dahl friction compensation gives the lowest sinusoidal position tracking error, outperforming the EKF based controller counterpart.