Abstract: Under the condition of high speed and high acceleration, the elastic property of ballscrew feed system will have great influence on its frequency response characteristics. And the frequency response characteristic can be identified through the analysis of amplitude and phase relationship between the torque input signal and acceleration output signal. The elastic ballscrew drives system is first modeled through lumped mass method, and the frequency response characteristic is calculated by Lagrange energy method and state space analysis. In the experimental identification process, in order to reduce the effects of quantization noise, white noise and color noise on identification signal, the Symlets wavelet filter method combined with finite differences is adopted to retrieve the acceleration signal that is severely contaminated by noise, then the amplitude and phase signals are extracted and the frequency response function is fitted. The simulation and experiment results demonstrate the Symlets wavelet can effectively reduce the adverse effect of noise signal and extract effective acceleration signal, so the overall accuracy of frequency response identification is improved.

Key words: Acceleration signal, Ballscrew, Frequency response, Servo driven system, Symlets wavelet filter