Online Estimation of Damping Coefficient Considering Variable Friction for Position Servo System

In the finite-time position tracking applications, the performance of the servo system is significantly affected by the friction. Meanwhile, the inevitable friction model parameters deviation and incomplete modeling will increase the friction estimation error, resulting in the higher position overshoot and longer settling time, especially for the light-load system. Damping control can mitigate the negative influence of the friction on the performance. However, the damping control with a fixed damping coefficient is not completely suitable for the system with variable friction characteristics, and the online tuning method of damping coefficient is not explicit and practical. Aiming at this issue, a modified online estimation method of damping coefficient is proposed in this article. First, the relationship between the damping coefficient and the differential of friction is derived, which allows to online tune the damping coefficient. Second, considering the difficulties in identifying the friction model parameters in orbit and the adaptability to different motors, the method for estimating the differential of friction and calculating the damping coefficient based on the observer is proposed. Finally, simulations and experiments for the motors with different friction characteristics are conducted, and the effectiveness and adaptability of the proposed method are validated by the results.