A Novel Modified Auto-regressive Moving Average Hysteresis Model

Abstract

A modified auto-regressive moving average (MARMA) model is proposed in this paper, which can be used to describe the dynamic hysteresis nonlinearity accurately. First, combined with the stability condition of auto-regressive moving average (ARMA) model, the Least Square approximation and the Lagrange Multiplier method (LSLM) are used to improve the traditional ARMA model. And then, according to the collected voltage-displacement data set, the parameters of the MARMA model are identified by LMLS method. Meanwhile, aiming at the difficulty of real-time displacement detection in the process of fast tool servo (FTS), a direct feedforward open-loop control (DFOC) strategy is designed based on the identified model. Finally, in order to verify the effectiveness and superiority of the method, a series of high frequency trajectory tracking and contrast experiments have been carried out successfully with the traditional PI and MARMA models. It shows that the MARMA model is nearly 20 times higher than the traditional PI model in terms of control accuracy and linearity, while the control bandwidth is achieved up to 200Hz.