1. INTRODUCTION
ISP, also known as gyroscope stabilization platform, is
widely used in vehicle, ship and airborne fire control
systems. It can effectively isolate the disturbance of the
carrier motion by servo control, so that the Line of Sight
(LOS) of the optical instrument carried by ISP can keep a
stable direction, or track and aim the target object according
to the orders. As an important equipment in the field of
modern military, the development of a stable platform for
high-precision and ultra-long-range target recognition has
become a key breakthrough point in the arms race of
countries around the world [1]. In the case of constant
hardware facilities, how to improve the system performance
through the control strategy is one of the hot spots in the
research of ISP. At present, the researchers put forward
various advanced ISP control methods, and some methods
have achieved good results, such as neural network control
[2], adaptive control [3], fuzzy control [4,5], optimal control
[6,7]. However, the implementation of these algorithms is
complex. In practical engineering, this task is mainly done
by PID. In addition, the real time environment of ISP is
relatively intricate, which is usually affected by parameter
uncertainty or external disturbances. Therefore, there is a
need to develop a controller which is versatile and
implemented easily.
In the control system, on the one hand, stronger is the
integral, better is tracking but worse is the stability of the
closed system. On the other hand, stronger is the differential,
higher is the stability margin and worse is the noise
rejection performance of the closed-loop system. The
FOPID with
format proposed by professor
Podlubny can be a proper trade-off controller between PID
controllers (i.e. ,,). In other words, FOPID can
well balance the accuracy and stability of the system. The
order λ and μ of
can be any real number, and the
traditional PID is the special case ofand .
Compared with PID, FOPID has the ability to improve the
capability of handling parametric uncertainties, disturbance
rejection and minimizing steady state errors. The above
characteristics inspire researchers to explore different
design methods. Luo et al. [10] designed the FOPID
controller by the constraints of gain crossover frequency,
phase margin and robustness in the frequency domain, and
simulation results show that the proposed controller
outperforms the PID controller. However, this method is
limited by the number of constraints, so only the POPI
controller or the FOPD controller can be designed. Meng et
al. [11] added two constraints on the basis of Luo, noise
suppression constraint and interference suppression
constraint. Compared with the PID controller, the proposed
FOPID controller significantly improves the control
performance, but the solution of five element nonlinear
equations is quite a difficult problem. It can be seen that
although the five parameters of FOPID improve the
flexibility of controller design, it increases the difficulty of
controller tuning. In the case of many parameters, the
traditional trial-and-error method not only takes a long time,
but also may not find the optimal parameters. One of the
tasks of this paper is to find a simple and effective FOPID
parameter tuning method.
IMC is a new control strategy based on the mathematical
model of the process. In this kind of controller, the plant
model is embedded in the controller. Compared with other
design schemes, the controller based on IMC is
mathematically very simple and has strong robustness to
external disturbances [12]. Previous researchers used the
IMC principle to design IMC-PID, and the designed
controller has the characteristics of simple tuning and easy
implementation. In recent years, it has been introduced into
the simplified design of FOPID. Saxena et al. [13] proposed
an IMC-FOPID controller for DC motor control. Different
from the traditional FOPID controller which needs five
tuning parameters, the controller has only two parameters,
which can be obtained by the expected gain crossover
frequency and phase margin. This method has achieved
good control effects in both simulation and hardware
Research and Design of Fractional-Order PID Controller for Inertial Stabilized
Platform via Internal Model Control
Yaqiong Zhang
1
, Jiandu Guo
1
, Yi Liu
1
, Jing Zhou
1
, Yu Sun
1
1. Information and Control Technology Department, China North Vehicle Research Institute, Beijing 100072
E-mail: zhyq@noveri.com.cn
Abstract: Inertial stabilized platform (ISP) is an important part of platform inertial navigation system, and its
mathematical model is established in this paper. In order to improve the accuracy and stability of the ISP, we propose a
Fractional-Order PID (FOPID) controller design method based on internal model control (IMC). Firstly, an IMC-PID
controller with only one parameter is constructed by the IMC principle. Then the FOPID controller is formed by adding
integrator order λ and differentiator order μ. Finally, these three parameters are easily turned by gain crossover frequency,
phase margin and gain robustness constraint. The Matlab results show that the proposed controller outperforms the PID
turned by the same frequency constraints method and can obtain good tracking performance, disturbance rejection and
robustness.
Key Words: Fractional-Order PID, Internal model control, Inertial stabilized platform