A hysteresis compensation method of piezoelectric actuator:
Model, identification and control
Changhai Ru
a,
, Liguo Chen
b
, Bing Shao
b
, Weibin Rong
b
, Lining Sun
b
a
College of Automation, Harbin Engineering University, Harbin 150001, People’s Republic of China
b
State Key Laboratory of Robotics and System, Harbin Institute of Technology (HIT), Harbin 150001, China
article info
Article history:
Received 17 September 2008
Accepted 25 April 2009
Available online 18 June 2009
Keywords:
Hysteresis
Adaptive inverse
Piezoelectric actuator
LMS
abstract
A major deficiency of piezoelectric actuators is that their open-loop control accuracy is seriously limited
by hysteresis. In this paper, a novel mathematical model is proposed to describe hysteresis precisely.
Based on the hysteresis model, an adaptive inverse control approach is presented for reducing
hysteresis. The weights of the m ain hysteresis loop are identified by using least mean square (LMS)
algorithm. The realization of an inverse feedforward controller for the linearization of a piezoelectric
actuator is formulated. Experiments were performed on a micro-positioning system driven by
piezoelectric actuators. The experimental results demonstrate that the positioning precision is
noticeably improved in open-loop operation compared to the conventional open-loop control without
any compensation.
& 2009 Elsevier Ltd. All rights reserved.
1. Introduction
Piezoelectric actuators are becoming increasingly impor-
tant in today’s positioning technology. These versatile micro-
positioning elements for movements are replacing traditional
positioning systems not only in research and development,
but also in industrial production such as scanning probe
microscopy (Yeh, Ni, & Panc, 2005) and adaptive optics system
(Fedrigo, Muradore, & Zilio, 2009). However, the piezoelectric
actuators have the disadvantage of hysteresis behavior, which
severely limits system performance such as giving rise to
undesirable inaccuracy or oscillations, even leading to instability.
Without the aid of further control technique to overcome this
problem, it is only possible to achieve a limited positioning
accuracy.
To guarantee a high precision positioning, there had
been tremendous amount of research efforts in the past to
reduce the effect of hysteresis. Newcomb and Flinn (1982)
suggested that there is an almost linear relation between the
displacement and charge, but the circuitry of this power driver is
complicated. Ge and Jouaneh (1995, 1997) and Schafer and
Janocha (1996) proposed a feedforward control technique
based on the Preisach model to compensate for the hysteresis.
A robust PID-control switching scheme is also proposed (Vagia,
Nikolakopoulos, & Tzes, 2008) to control a micro-actuator.
Recently, Prandtle–Ishlinskii operator is used to control
piezoelectric actuator proposed by Kuhnen and Janocha (1998,
1999, 2000). Tao and Kokotovic (1995) and Webb, Kurdila
and Lagoudas (2000) proposed an adaptive hysteresis inverse
cascade with the system, so that the system becomes a linear
structure with uncertainties. Moreover, there are many
recent works about piezoelectric actuator control and application
in industry, such as the vibratory feeding (Hu & Farson, 2007), the
piezoforce sensor (Kim & Yim, 2008), high-pressure injection
control (Park, Kim, & Lee, 2006) and precise positioning
(Liaw, Shirinzadeh, & Smith, 2008; Lin, Shieh, & Huang, 2008;
Shen, Jywe, & Shu, 2008). An adaptive model-based predictive
method is designed for high-frequency waveform tracking
by a PZT bimorph actuator in industrial vibratory feeding (Hu &
Farson, 2008). Another adaptive controller is also proposed to
address the non-linearities and uncertainties of piezoelectric
actuator (Putra, Huang, Tan, Panda, & Lee, 2007). The present
paper describes an alternative solution for the compensation of
the hysteresis of piezoelectric stack actuator. A mathematical
model is introduced, which can accurately describe the hysteresis
phenomenon. The main hysteresis referenced curves are des-
cribed by the weighted superposition of many Prandtle–Ishlinskii
hysteresis operators, which is feasible to realize an online
adaptive control scheme. Then hysteresis are derived by the
identification of least mean square (LMS) adaptive algorithm.
Combining this inverse model with adaptive control techniques,
an intelligent control algorithm is developed. As a result, the effect
of the hysteresis is reduced obviously and the desired control
accuracy is guaranteed.
ARTICLE IN PRESS
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locat e/conengprac
Control Engineering Practice
0967-0661/$ - see front matter & 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.conengprac.2009.04.013
Corresponding author. Tel./fax: +86 45182518806.
E-mail address: rchhai@yahoo.com.cn (C. Ru).
Control Engineering Practice 17 (2009) 1107–1114