Disturbance Rejection control for Discrete Time-delay Nonlinear System
Ma Hui
1
, Tang Gong-You
1
, Hu Wei
2
, Zhang Bao-Lin
3
,Su Hao
1
1. College of information science and technology, Ocean University of China, Qingdao 266100, China
E-mail: {mahui@ouc.edu.cn, gtang@ouc.edu.cn, asfreedom@163.com}
2. North China Sea Marine Forecast Center, State Oceanic Administration, Qingdao 266033, China
E-mail: huwei@bhfj.gov.cn
3. College of Sciences, China Jiliang University, Hangzhou, 310018, China
E-mail: zhangbl@cjlu.edu.cn
Abstract: This paper studies disturbance rejection problem for discrete time-delay nonlinear systems. The optimal control of
original systems causes a nonlinear Two Points Boundary Value (TPBV) problem with both time-delay terms and time-advanced
terms. By using of Successive Approximation Approach (SAA), the original nonlinear TPBV problem is transformed into a
sequence of nonhomogeneous linear TPBV problems without time-delay terms and time-advanced terms. We proved that the
sequence of solutions uniformly converges to the optimal disturbance rejection control law of the original systems through two
lemmas. By finite iterations of the solution sequence, the iterative procedure can be stopped and suboptimal control law is
obtained. Based on reduced-order disturbance observer, the presented control law can be easily completed and physically
realizable. Simulation results of two cases verify the validity and effectiveness of the presented controller.
Key Words: Nonlinear, time-delay, disturbance observer, suboptimal, Successive Approximation Approach
1 Introduction
In real world, all control systems have the nonlinear
characteristic, and the performances are often affected by
time delay and outside disturbances. And the time delays
and disturbances should be considered in the design of the
controller.
There are many valuable research results on the nonlinear
disturbance rejection. For the uncertain systems, [1]
presents an output feedback control based on the nonlinear
disturbance observer. And robust control is adapted to
depress the disturbances of nonlinear systems[2, 3]. [4-5]
research optimal disturbance rejection based on disturbance
observer, and make applications in the vibration control of
offshore platforms. Adaptive controllers are also presented
in nonlinear systems [6,7].
The delays in data collection and transmission, computer
and application of control are inevitable. The time delay
may cause the decrease of system performance, and even
destabilize the system. In the literatures, there are many
types of controller designed for systems with delay, for
example, feedback control [8], feedforward control based
on nonlinear Smith predictor [9], predictive control [10],
HFRQWURO [11] , disturbance estimator and observer[12,13],
and so on. In some researches, system model transmission
method [14] transfers the time delay systems into systems
without delay, and designs the vibration control law for
offshore platform based on the transferred system. The time
delay can be introduced into controller artificially and the
proper delay may benefit the performance of the system. [15,
16] propose delayed controllers to reduce the oscillation
amplitudes of the offshore steel jacket platform.
The optimal disturbance rejection for time delay
nonlinear system is very important in the literature by now.
*
This work is supported by National Natural Science Foundation
(NNSF) of China under Grant 41276085 ,61379029 and 61572448.
Especially the optimal disturbance rejection problem for
discrete time nonlinear delay systems is meaningful and
important. Considering the complexity of system model and
computation procedure of the optimal control, it is easier to
obtain suboptimal control in real world. In this work we
study the suboptimal disturbance rejection problem of time
delay nonlinear system based on the optimal control law by
Successive Approximation Approach (SAA) [17, 18].
The rest of paper is organized in the followings.
Section 2 describes the problem and gives the lemmas for
the convergence of solutions. Section 3 solves the nonlinear
TPBV problem with time advance and time delay terms, and
designs the controller by Successive Approximation
Approach (SAA). An iterative process is constructed, which
is uniformly convergent to the original optimal control
problem. By infinite iterations, we get suboptimal control of
original system. Based on the disturbance observer with
reduced dimension, we can design physically realizable
controller. In section 4, simulation examples in two cases
are given to verify the controller. Solutions are obtained in
section 5.
2 Problem Statement
Consider the following discrete time-delay nonlinear
systems with disturbance
( 1) ((), ()) ( ) (),
0,1, ,
() () [ ,0]
kkkkhk
kN
kkkh
1
xHxuAxDv
x ij
"
(1)
where
n
x \
is state variable,
m
u \
is control,
r
v \
is
persistent outside disturbance,
ij
is initial value of state
variable,
1
A
and
D
are constant matrices,
0h !
is time
delay,
1
:( )
nm n
CR R U RuoH
is nonlinear term,
satisfying
(0,0) H0
.
Proceedings of the 35th Chinese Control Conference
Jul
27-29, 2016, Chen
du, China
2571