Friction Compensation for Accurate Positioning in DC Drive Tracking System

Friction is one of the greatest obstacles in high precision positioning systems. It can cause steady state errors and tracking errors, while it may results in limit cycles. Therefore, the influence of friction on the response of systems, such as a DC servo drive, must be seriously considered. In this paper, considering the LuGre model of friction as part of the process taken into account, a hybrid structure of a feedback compensation scheme is proposed. A compensation signal is provided by a model-based observer via an inverse model of DC motor. The model-based observer is tuned using a load disturbance torque estimator. Thereafter, the whole structure is implemented in a typical position closed-loop in which the feedback consists of a state-feedback (P) part and the model-based compensation term. Good transient responses and tracking steady-state errors are achieved for different micrometer scale reference test signals.