Dynamic Parameters Identification for the Feeding System of CommercialNumerical Control Machine

摘要

The precision of high-speed CNC (Computer Numerical Control) machine are greatly influenced by the dynamiccharacteristics of servo system. To establish the servo system model accurately, the internal signals of NC, e.g. motorcurrent and rotate speed were adopt as the input or output signal for the system. The ARMA linear identification modelwas also established to identify the dynamic parameters of the mechanical parts in the servo system, including equivalentinertia, equivalent damping and so on. The friction Stribeck curve was obtained by necessary experiments. The nonlinearfriction model was linearized by using higher-order Taylor expansion, and the five parameters of the Stribeck frictionmodel were identified. To verify the effectiveness of the closed-loop identification, experiments are carried out on thefeeding system of a commercial NC machine. Signals of servo motor current and rotating rate which are offered in manymodern CNC machine tools are needed for the identification and experiments results show that the proposed method performswell with rapid convergence and accurate results and parameters of Stribeck model can be obtained accurately byidentification. The method is suited for industrial condition and has its practicality.