Design mechanism of Sampling Frequency on Mechanical Parameter Identification in A Two-Mass Servo Drive System

In servo drive systems, there usually exist elastic components that inevitably cause mechanical oscillations. In order to achieve self-tuning characteristics of the controller and improve the control performance, establishing an accurate model is crucial. In this paper, the recursive least square (RLS) method is first applied for the identification of mechanical parameters for the two-mass elastic system. In order to improve the accuracy and robustness of the identification algorithm, the derivation of the mechanical oscillation from the servo system is then investigated. Last, this paper explores the influence of sampling frequency on identification accuracy. By analyzing the Shannon Sampling Theorem, the oscillation theory of the servo drive system, the identification algorithm, and also the practical engineering factors, the suitable design mechanism of sampling frequency is deduced to get a favorable identification results.