Friction-induced disturbances degrade the accuracy of positioning systems. Although the LuGre model balances physical fidelity and parameter parsimony, its identification remains challenging due to nonlinear dynamics and parameter coupling. This paper proposes a systematic LuGre friction identification methodology combining a gradient-based optimization algorithm with a friction-guided experimental design. The approach relies on a single-objective cost function and a two-step parameter initialization strategy. The methodology is validated through simulation and experimental results obtained on an MKS Instruments motorized linear stage, illustrating its suitability for accurate friction modeling in industrial positioning applications.
Urquiza et al. (Wed,) studied this question.