Feedforward Tracking Controller Design Based on the Identification of Low Frequency Dynamics

Abstract

Several methodologies are proposed for identifying the dynamics of a machine tool feed drive system in the low frequency region. An accurate identification is necessary for the design of a feedforward tracking controller, which achieves unity gain and zero phase shift for the overall system in the relevant frequency band. In machine tools and other mechanical systems, the spectrum of the reference trajectory is composed of low frequency signals. Standard least squares fits are shown to heavily penalize high frequency misfit. Linear models described by the output-error (OE) and Autoregressive Moving Average with eXogenous Input (ARMAX) models display better closeness-of-fit properties at low frequency. Based on the identification, a feedforward compensator is designed using the Zero Phase Error Tracking Controller (ZPETC). The feedforward compensator is experimentally shown to achieve near-perfect tracking and contouring of high-speed trajectories on a machining center X-Y bed.