International Journal of Trend in Scientific

Research and Development (IJTSRD)

International Open Access Journal
ISSN No: 2456 - 6470 | www.ijtsrd.com | Volume - 2 | Issue – 1

Chaos Suppression and Practical Stabilization of Uncertain

Generalized Duffing
Duffing-Holmes
Holmes Control Systems
with Unknown Actuator Nonlinearity

Yeong-Jeu Sun
Professor, Department of Electrical Engineering,
I-Shou
Shou University
University, Kaohsiung, Taiwan

ABSTRACT
In this
his paper, the concept of practical stabilization for
nonlinear systems is introduced and the practical
stabilization of uncertain generalized Duffing
Duffing-Holmes
control systems with unknown actuator nonlinearity is
explored. Based on the time-domain
domain approach with
differential inequalities, a single control is presented
such that the practical stabilization for a class of
uncertain generalized Duffing-Holmes
Holmes systems with
unknown actuator nonlinearity can be achieved.
Moreover, both of the guaranteed exponential
convergence rate and convergence radius can be
correctly calculated Finally, some numerical
simulations are given to demonstrate the feasibility and
effectiveness of the obtained results.
Keywords: Practical synchronization, chaotic
haotic system,
uncertain generalized Duffing-Holmes
Holmes systems,
unknown actuator nonlinearity, chaos
haos suppression
1. INTRODUCTION
Chaotic dynamic systems have been extensively
investigated in past decades; see, for instance, [1
[1-13]
and the references therein. Very often, chaos in man
many
dynamic systems is an origin of instability and an
origin of the generation of oscillation. Generally
speaking, the robust stabilization of uncertain dynamic
systems with a single controller is in general not as
simple as that with multiple controllers. In the past
decades, various methodologies in robust control of
chaotic system have been offered, such as control
approach, adaptive control approach, variable structure
control approach, backstepping control approach,
adaptive sliding mode control approach,
appro and others.
In this paper, the concept of practical stabilizability for
uncertain dynamic systems is introduced and the
practical stabilizability of uncertain generalized
Duffing-Holmes
Holmes control systems with unknown
actuator nonlinearity will be studied.
stud Using the time-
domain approach with differential inequality, a single
control will be designed such that the practical
stabilization can be achieved for a class of uncertain
generalized Duffing-Holmes
Holmes systems with unknown
actuator nonlinearity. Not only ly the convergence radius
and guaranteed exponential convergence rate can be
arbitrarily pre-specified,
specified, but also the unknown actuator
nonlinearity and mixed uncertainties can be
simultaneously overcome by the proposed single
control. Several numerical simulations
simu will also be
provided to illustrate the use of the main results.
The rest of paper is organized as follows. The problem
formulation and main results are presented in Section 2.
Numerical simulations are given in Section 3 to
illustrate the effectiveness
veness of the developed results.
Finally, some conclusions are drawn in Section 4.

@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 2 | Issue – 1 | Nov-Dec

Dec 2017 Page: 1285
 International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
2. PROBLEM FORMULATION AND MAIN t d 2t
0 dt e  st dt  e  st   s0
2t

RESULTS
t

Nomenclature   e 2t  a dt
0

 n
the n-dimensional real space 
2

a 2 t

e  1 ,  t  0.

a the modulus of a complex number a

Consequently, one has
I the unit matrix
 a   2t a
s t    s 0  e  , t 0.
 2  2
A T
the transport of the matrix A

x the Euclidean norm of the vector x  n This completes the proof. □

In this paper, we consider the following uncertain

min ( P ) the minimum eigenvalue of the matrix P with
generalized Duffing-Holmes control systems with
real eigenvalues uncertain actuator nonlinearity described as
 (A) the spectrum of the matrix A x1  x2 , (1a)
P0 the matrix P is a symmetric positive definite
x2  q1 x1  q2 x2  q3 x13  q4 x15
matrix  q5 cosw1t   f x1 , x2  (1b)
  u ,  t  0.
Before presenting the problem formulation, let us
introduce a lemma which will be used in the proof of
where x  x1 x2 T   21 is the state vector, u   is the
the main theorem.
input, f x1 , x2  represents the mixed uncertainties
Lemma 1 (unmodeled dynamics, parameter mismatchings,
external excitations, and disturbance), and  u 
If a continuously differentiable real function st  represents the uncertain actuator nonlinearity. For the
satisfies the inequality existence of the solutions of (1), we assume that the
unknown terms f x1 , x2  and  u  are all continuous
st   a  2 st ,  t  0 , functions. It is well known that the system (1) without
any uncertainties (i.e., f x1 , x2    u   0 ) displays
with a  0 and   0 , then chaotic behavior for certain values of the parameters
[1]. In this paper, the concept of practical stabilization
 a  2t a
s t    s 0    e  , t 0. will be introduced. Motivated by time-domain approach
 2  2 with differential inequality, a suitable control strategy
will be established. Our goal is to design a single
Proof. It is easy to see that control such that the practical stabilization for a class of
uncertain nonlinear systems of (1) can be achieved.
e 2t  st   e 2t  2 s t 


d 2t
dt
 
e  s t   e 2t  a,  t  0. Throughout this paper, the following assumption is
made:
It results that (A1) There exist continuous function f x1 , x2   0 and
positive number r1 such that, for all arguments,

f x1 , x2   f  x1 , x2  ,

u   u   r1u 2 .

@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 2 | Issue – 1 | Nov-Dec 2017 Page: 1286
 International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
Remark 1 ht  : q1 x1  q2 x2  q3 x13  q4 x15
 q5 cosw1t     1 x1 (4)
2
Generally speaking, if the uncertain actuator
nonlinearity satisfies  2  1x2  f ,

r1  u 2  u   u   r2  u 2 ,  u   ,  p1 p3 
where P :   0 is the unique solution to the
 p3 p2 
we often ascribe r2 as the gain border and r1 as the gain following Lyapunov equation
reduction endurance. Thus, we know r2   from (A1);
T
see Figure 1.   1 
   12    2 P
  (5)
A precise definition of the practical stabilization is   1 
 P   2 I ,
given as follows, which will be used in subsequent     12
   2 
main results.
with   0 . In this case, the guaranteed convergence
Definition 1 radius and exponential convergence rate are  and  ,
The uncertain system (1) is said to realize the practical respectively.
stabilization, provided that, for any   0 and   0 , Proof. From (1), the state equation can be represented
there exist a control u :  ,   such that the state as
trajectory satisfies
 0 1 
x   x
xt     e t   ,  t  0,     12
 2  2

for some   0 . In this case, the positive number  is

0 

      f  q1 x1  q2 x2  q3 x13  q4 x15
1
called the convergence radius and the positive number
 q5 cosw1t     1 x1  2  2x2
2

 is called the exponential convergence rate. Namely,

 Ax  B    f  q1 x1  q2 x2  q x 3
3 1
the practical stabilization means that the states of
 q x  q5 cosw1t     1 x1
5 2

system (1) can converge to the equilibrium point at 4 1

 2  2 x2 ,  t  0,
x  0 , with any pre-specified convergence radius and
exponential convergence rate. Obviously, a control  0 1  0 
where A :   and B :   . Clearly, one
   1  2  2
system, having small convergence radius and large 2
1
exponential convergence rate, has better steady-state has   A     1 . This implies A  I is Hurwitz and
response and transient response.
the Lyapunov equation of (5) has the unique posive
Now we present the main result for the practical definite solution P. Let
stabilization of uncertain systems (1) via time-domain
V  xt   xT t Pxt  . (6)
approach with differential inequalities.

Theorem 1 The time derivative of V xt  along the trajectories of

the system (1) with (2)-(6) is given by
The uncertain systems (1) with (A1) realize the
practical stabilization under the following control 
V xt   xT AT P  PA x  2 xT PB

   f  q1 x1  q2 x2  q3 x13  q4 x15
u t   r t    p3 x1  p2 x2  , (2)
 q5 cosw1t     1 x1  2  2 x2
2

r t  :
h 2 t  ,(3)
r1  ht   p3 x1  p2 x2    r1    min P 
2

@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 2 | Issue – 1 | Nov-Dec 2017 Page: 1287
 International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
 xT  2P  2 I  x  2 xT PB   Consequently, we conclude that

 2 xT PB q1 x1  q2 x2  q3 x13  q4 x15
V x0    2 min P 
 q5 cosw1t     1 x1  2  2 x2
2 xt   e  2t   
2

  min  P  
 f
V x0    2 min P 
 e  2t   
2

 2xT Px  2 xT PB    2ht   xT PB   min  P  

 V  x0   2 min P   t

2   e
 2xT Px     u     2h  xT PB 
min P 
r  
  ,  t  0.
2
 
 2V     r1u 2  2h  x T PB
This completes the proof. □
r

2 3. NUMERICAL SIMULATIONS
 2V  2r1  r  x T PB  2h  x T PB
Consider the following uncertain generalized Duffing-
2
 2V  2r1  r  p3 x1  p2 x2 Holmes control systems with unknown actuator
 2h  p3 x1  p2 x2 nonlinearity described as

 2V  x1  x2 , (8a)

2
2r1  p3 x1  p2 x2  h 2
x2  q1 x1  q2 x2  q3 x13  q4 x15
r1  h  p3 x1  p2 x2    r1   2  min P 
 q5 cosw1t   f x1 , x2  (8b)
 2h  p3 x1  p2 x2
  u ,  t  0,
 2V
where
2h  p3 x1  p2 x2    r1   2  min P 

r1  h  p3 x1  p2 x2    r1   2  min P  q1  1 , q2  0.25 , q3  1 , q4  0 , q5  0.3 , w1  1 ,
2
 2V   
 r1  f x1 , x2   a  x12 ,  u   b  u  c  u 3 ,
 r1  h  p3 x1  p2 x2    r1   2  min P  
 , By the 1  a  1, 1  b  3, c  0.
 r1  h  p3 x1  p2 x2    r1    min P  
2

 t  0. Our objective, in this example, is to design a feedback

inequality control such that the uncertain systems (8) realize the
practical stabilization with the exponential convergence
 yz 
x   xz,  x  0, y  0, and z  0 . rate   3 and the convergence radius   0.1 . The
 yz condition (A1) is clearly satisfied if we let f x1 , x2   x12 ,
It can be deduced that r1  1 . From (5) with   3 , we have

V xt   2V xt   141 26.5

(7) P , min P   0.5,
 2    min P ,  t  0.
2
26.5 5.5 

Thus, from Lemma 1, (6), and (7), it can be readily p2  5.5, p3  26.5.

obtained that
From (4), it can be readily obtained that
min P  xt   V xt 
2

ht  : x1  0.25 x2  x13  0.3 cos t  16 x1  8 x2

e  2 t
V x0   2

min P 
 x12 .
  min P ,  t  0.
2

@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 2 | Issue – 1 | Nov-Dec 2017 Page: 1288
 International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
From (3) with   0.1 , one has REFERENCES
h 2 t 
r t  : . Finally,
ht   26.5 x1  5.5 x2  0.015
control, given by (2), can now be calculated as
u t   r t   26.5 x1  5.5 x2  . (9)
Consequently, by Theorem 1, we conclude that system
(8) with the control (9) is practically stable, with the
exponential convergence rate   3 and the guaranteed
convergence radius   0.1 . The typical state trajectories
of uncontrolled systems and controlled systems are
depicted in Figure 2 and Figure 3, respectively.
Besides, the time response of the control signal is
depicted in Figure 4. From the foregoing simulations
results, it is seen that the uncertain dynamic systems of
(8) achieves the practical stabilization under the control
law of (9).
4. CONCLUSION
In this paper, the concept of practical stabilization for
nonlinear systems has been introduced and the practical
stabilization of uncertain generalized Duffing-Holmes
control systems with unknown actuator nonlinearity has
been studied. Based on the time-domain approach with
differential inequalities, a single control has been
presented such that the practical stabilization for a class
of uncertain generalized Duffing-Holmes systems with
unknown actuator nonlinearity can be achieved. Not
only the unknown actuator nonlinearity and mixed
uncertainties can be simultaneously overcome by the
proposed control, but also the convergence radius and
guaranteed exponential convergence rate can be
arbitrarily pre-specified. Finally, some numerical
simulations have been offered to show the feasibility
and effectiveness of the obtained results.
ACKNOWLEDGEMENT
The author thanks the Ministry of Science and
Technology of Republic of China for supporting this
work under grants MOST 105-2221-E-214-025, MOST
106-2221-E-214-007, and MOST 106-2813-C-214-
025-E.
@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 2 | Issue – 1 | Nov-Dec 2017
the desired 1) X. Liu, L. Bo, Y. Liu, Y. Zhao, Y. Zhao, J. Zhang,
N. Hu, S. Fu, M. Deng, “Detection of micro-cracks
using nonlinear lamb waves based on the
Generalized Generalized Duffing-Holmes system,”
Journal of Sound and Vibration, vol. 405, pp. 175-
186, 2017.
2) V. Wiggers and P.C. Rech, “Multistability and
organization of periodicity in a Van der Pol-Duffing
oscillator,” Chaos, Solitons & Fractals, vol. 103, pp.
632-637, 2017.
3) H. Shi and W. Li, “Research on weak resonance
signal detection method based on Duffing
oscillator,” Procedia Computer Science, vol. 107,
pp. 460-465, 2017.
4) J. Niu, Y. Shen, S. Yang, and S. Li, “Analysis of
Duffing oscillator with time-delayed fractional-
order PID controller,” International Journal of Non-
Linear Mechanics, vol. 92, pp. 66-75, 2017.
5) J. Kengne, Z.N. Tabekoueng, and H.B. Fotsin,
“Coexistence of multiple attractors and crisis route
to chaos in autonomous third order Generalized
Generalized Duffing-Holmes type chaotic
oscillators,” Communications in Nonlinear Science
and Numerical Simulation, vol. 36, pp. 29-44, 2016.
6) J. Li, M. Mao, and Y. Zhang, “Simpler ZD-
achieving controller for chaotic systems
synchronization with parameter perturbation, model
uncertainty and external disturbance as compared
with other controllers,” Optik-International Journal
for Light and Electron Optics, vol. 131, pp. 364-
373, 2017.
7) H. Tirandaz and A.K. Mollaee, “Modified function
projective feedback control for time-delay chaotic
Liu system synchronization and its application to
secure image transmission,” Optik-International
Journal for Light and Electron Optics, vol. 147, pp.
187-196, 2017.
8) H.P. Ren, C. Bai, Q. Kong, M.S. and Baptista, C.
Grebogi, “A chaotic spread spectrum system for
underwater acoustic communication,” Physica A:
Statistical Mechanics and its Applications, vol. 478,
pp. 77-92, 2017.
9) K.B. Deng, R.X. Wang, C.L. Li, and Y.Q. Fan,
“Tracking control for a ten-ring chaotic system with
an exponential nonlinear term,” Optik-International
Journal for Light and Electron Optics, vol. 130, pp.
576-583, 2017.
Page: 1289
 International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
10) E.M. Bollt, “Regularized forecasting of chaotic 1

dynamical systems,” Chaos, Solitons & Fractals,

vol. 94, pp. 8-15, 2017. 0.5
x1

11) M.J. Mahmoodabadi, R.A. Maafi, and M. 0

Taherkhorsandi, “An optimal adaptive robust PID

x 1(t); x 2(t)
-0.5
controller subject to fuzzy rules and sliding modes
for MIMO uncertain chaotic systems,” Applied Soft -1

Computing, vol. 52, pp. 1191-1199, 2017. x2

-1.5

12) W. Jiang, H. Wang, J. Lu, G. Cai, and W. Qin,

“Synchronization for chaotic systems via mixed- -2
0 0.5 1 1.5 2 2.5 3

objective dynamic output feedback robust model t (sec)

predictive control,” Journal of the Franklin Institute,

vol. 354, pp. 4838-4860, 2017. Figure 3: Typical state trajectories of the feedback-
controlled system of (8) with (9).
13) G. Xu, J. Xu, C. Xiu, F. Liu, and Y. Zang, “Secure
communication based on the synchronous control of
hysteretic chaotic neuron,” Neurocomputing, vol.
227, pp. 108-112, 2017.

 u 

r1u
u

Figure 1: Uncertain actuator nonlinearity.

2
x1: the Blue Curve x2: the Green Curve
1.5

0.5
x 1(t); x 2(t)

-0.5

-1

-1.5

-2
0 50 100 150 200 250 300 350 400 450 500
t (sec)

Figure 2: Typical state trajectories of the uncontrolled

system of (8).

@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 2 | Issue – 1 | Nov-Dec 2017 Page: 1290