Kybernetika 51-2015-1 10
Kybernetika 51-2015-1 10
Kybernetika 51-2015-1 10
Terms of use:
© Institute of Information Theory and Automation AS CR, 2015
Institute of Mathematics of the Czech Academy of Sciences provides access to digitized documents
strictly for personal use. Each copy of any part of this document must contain these Terms of use.
This document has been digitized, optimized for electronic delivery and stamped
with digital signature within the project DML-CZ: The Czech Digital Mathematics
Library http://dml.cz
KYBERNETIKA — VOLUME 51 (2015), NUMBER 1, PAGES 137–149
In this paper, we investigate the finite-time stochastic synchronization problem of two chaotic
systems with noise perturbation. We propose new adaptive controllers, with which we can
synchronize two chaotic systems in finite time. Sufficient conditions for the finite-time stochastic
synchronization are derived based on the finite-time stability theory of stochastic differential
equations. Finally, some numerical examples are examined to demonstrate the effectiveness
and feasibility of the theoretical results.
1. INTRODUCTION
Christian Huygens firstly picked up the appearance of synchronization about 350 years
ago. In fact, the significance of synchronization, especially for chaotic systems, was not
completely achieved until Pecora and Carroll published their pioneering work on chaos
synchronization in 1990 [31]. After that, a large variety of synchronization phenomena, in
many chaotic systems [7] and dynamical networks [12, 18], have been widely investigated
in different areas including physics, chemistry, biology, etc [28]. And lots of important
real applications have been found in many fields, such as information processing, hori-
zontal platform systems [13], secure communication [4, 5], biological system [15], rotating
pendulums [14], control processing, chemical reactions and so on [6, 10, 20, 30, 32]. As
we all known, a focused problem, in chaos synchronization, is to make the states of the
slave system follow the master system with an appropriate controller. Due to a wide
variety of applications, many approaches and controllers have been presented, including
adaptive control [17, 21, 25], optimal control [9, 34], sliding mode control [33], delayed
Lur’e systems control [19], the open-loop-closed-loop coupling technology [16], linearly
coupled ordinary differential systems analysis [27] and so on.
Recently, the finite-time synchronization of two chaotic systems has been investigated
by many researchers [1, 2, 3, 8, 24, 26, 35, 37, 40]. Finite-time generalized synchroniza-
tion of chaotic systems with different order has been studied in Ref. [8]. The adaptive
feedback controller was proposed to realize the finite-time synchronization for a class
DOI: 10.14736/kyb-2015-1-0137
138 J. WU, Z. C. MA, Y. Z. SUN AND F. LIU
of chaotic and hyperchaotic systems [35]. In Ref. [1], a robust adaptive controller was
introduced to realize finite-time chaos synchronization between two different chaotic
systems in the presence of model uncertainties, external disturbances, fully unknown
parameters, and input nonlinearities. In Ref. [26], the author implemented and tested
experimentally a four-dimensional hyperchaotic system and investigated the synchro-
nization of the system in a finite time, based on the finite-time stability theory. In Ref.
[37], the authors investigated the global finite-time synchronization of a class of second-
order nonautonomous chaotic systems via a master-slave coupling. In Ref. [40], the
adaptive finite-time synchronization of different coupled chaotic systems with unknown
parameters was explored.
However, the problems of finite-time synchronization in Refs. [1, 8, 26, 35, 37, 40] did
not take noise perturbation into consideration. It deserves pointing out that noise per-
turbation is widespread in both natural and artificial systems. For instance, because the
atmospheric effects and processes such as cloud cover, pollution, etc., are seasonal and
stochastic in nature, sunshine duration and solar irradiation are modeled in a stochastic
way. Therefore, it has more practical value to explore the influence of circumstance
noise on the finite-time synchronization of chaotic systems. The main contribution of
this paper is to propose an adaptive feedback controller, which can realize the finite-time
stochastic synchronization between two chaotic systems with noise perturbation. Based
on the finite-time stability theory for stochastic differential equations, sufficient condi-
tions for the finite-time stochastic synchronization are obtained. Finally, some numerical
examples are examined to illustrate the effectiveness of the analytical results. The ef-
fects of control parameters and noise intensity on the convergence time are numerically
demonstrated.
The rest of this paper is organized as follows. In Section 2, the problem statement
and some useful preliminaries are given. Based on the stability theory of stochastic
differential equations, sufficient conditions for the finite-time stochastic synchronization
are derived analytically in Section 3. In Section 4, some numerical examples are given
to show the effectiveness of the theoretical results. Finally, some conclusions are drawn
in Section 5.
Assumption 2.1. For function f (x) there exists a nonnegative constant l satisfying
For the noise intensity function, because the speed of the environmental fluctuations
is far less than the change rate of practical systems, we have the following assumption:
Assumption 2.2. The noise intensity function σi (ei (t)) (i = 1, 2, . . . , n) satisfies the
Lipschitz condition and there exists a nonnegative constant q such that
Moreover, σ(0) ≡ 0.
Definition 2.3. (Yin et al. [38]) The trivial solution of (4) is said to be finite-time
stable in probability, if the equation admits a unique solution for any initial data x(0) ∈
Rn , denoted by x(t; x(0)), moreover, the following statements hold:
(i) For every pair of ε ∈ (0, 1) and r > 0, there exists a δ = δ(ε, r) > 0 such that
Definition 2.4. Systems (1) and (2) are said to achieve the finite-time stochastic syn-
chronization if, for any initial states xi (0), yi (0) ∈ Rn \ {0}, there exists a finite time
function T0 such that
where T0 = inf{T : xi (t; xi (0)) = yi (t; yi (0)), ∀ t ≥ T } is called the stochastic setting
time.
Lemma 2.5. (Yin et al. [39]) For system (4), define T0 (x0 ) = inf{T ≥ 0 : x(t; x0 ) =
y(t; y0 ), ∀ t ≥ T }. Assume that system (4) has the unique global solution. If there exists
a positive definite, twice continuously differentiable and radially unbounded Lyapunov
function V : Rn → R+ , K∞ class functions µ1 and µ2 , positive real numbers c > 0 and
0 < γ < 1, such that for all x ∈ Rn and t ≥ 0,
Theorem 3.1. Suppose that the Assumptions 2.1 and 2.2 hold and there exist a suf-
ficiently large positive constant L satisfying L ≥ l + q, then systems (1) and (2) can
achieve finite-time stochastic synchronization under the following adaptive controllers:
ki + k̄
ui (t) = εi ei + ki sign(ei ) − sign(ei ), if ei 6= 0,
|ei |
ui (t) = 0, if ei = 0, (7)
P r o o f . From Eqs. (1) and (2), we can get the error system as follows:
Therefore, under Assumptions 2.1 and 2.2, it is from the theory of stochastic differential
equation that the error system (9) possesses a unique global solution on t ≥ 0, denoted by
ei (t, ei (0)) for any initial data ei (0) = yi (0) − xi (0). And ei (t, 0) ≡ 0 is a trivial solution
of the error dynamics (9). Obviously, if this trivial solution is globally stochastically
finite-time stable, then the finite-time stochastic synchronization between systems (1)
and (2) could be realized for every initial data.
Consider the following Lyapunov function:
n n n
1X 2 1X 1X
V = ei + (ki + k̄)2 + (εi + L)2 . (10)
2 i=1 2 i=1 2 i=1
Thus the diffusion operator L defined in (5) onto the function V along the error system
(9) gives
n
X
LV (t) = ei (t)[f (yi ) − f (xi ) + εi ei (t) + ki sign(ei )
i=1
n
ki + k̄ 1X
− sign(ei )] + trace(σiT σi )
|ei | 2 i=1
n
X n
X
+ (ki + k̄)k̇i + (εi + L)ε̇i . (11)
i=1 i=1
Substituting k̇i and ε̇i into the rihgt-hand side of Eq. (11), we have
n
X n
X n
X
LV (t) = ei (t)[f (yi ) − f (xi )] + εi e2i + ki |ei |
i=1 i=1 i=1
n n
X 1X
− (ki + k̄) + trace(σiT σi )
i=1
2 i=1
n
X n
X
− (ki + k̄)|ei | − (εi + L)e2i . (12)
i=1 i=1
we obtain
n
hX n
X i 21
LV (t) ≤ − e2i + (ki + k̄)2
i=1 i=1
1
, −(2V1 ) 2 ,
P
n Pn
where V1 = 12 2
i=1 ei +
2
i=1 (ki + k̄) .
Note that LV ≤ 0. Then
E V̇ = LV ≤ 0.
Thus, V is non-increasing in mean square. Then, there exists a upper bound V ∗ , such
that
V1 ≤ V ≤ V ∗ .
V1
Let θ = V∗ ≤ 1, then
θV ≤ θV ∗ = V1 .
Thus, we have
√ 1
LV (t) ≤ − 2θV 2 . (16)
According to Lemma 2.5, the trivial solution of the error system (9) is globally stochas-
tically finite-time stable. This means that the synchronization between systems (1) and
(2) could be achieved in finite time for almost every initial data, and the finite time is
estimated by
r
2 1
E[T0 ] ≤ T1 = V 2 (0), (17)
θ
Pn Pn i
where V (0) = [ i=1 ei (0)2 + i=1 (ki (0) + k̄)2 /2. This completes the proof.
Remark 3.2. From the inequality (14) we can see that, for any high level noise, there
exits a sufficiently large positive constant L such that the finite-time stochastic syn-
chronization is realized in probability. Hence the synchronization is robust to the noise
Finite-time synchronization of chaotic systems 143
perturbation. The convergence time of proposed algorithm is closely related to the vi-
olation of noise intensities. From the inequality (14) and the Itô formula, one can also
see that for fixed V (0), the synchronization time is inversely proportionally to the noise
intensity.
If system (4) is free of noise perturbation, namely σi (ei (t)) ≡ 0(i = 1, 2, . . . , n), from
Theorem 3.1, we have the following corollary:
Corollary 3.3. Let Assumption 2.1 holds. If σi (ei (t)) ≡ 0 in system (2) and L ≥ l ,
then the chaotic systems (1) and (2) can achieve finite-time synchronization under the
following control scheme:
ki + k̄
ui (t) = εi ei + ki sign(ei ) − sign(ei ), ei 6= 0, i = 1, 2, . . . , n.
|ei |
For ei 6= 0, the feedback gains εi and ki are adapted according to the following updated
laws ε̇i (t) = −e2i (t), k̇i (t) = −|ei (t)|; For ei = 0, we set ki ≡ −k̄ and εi ≡ −L.
4. SIMULATION RESULTS
In this section, a three-dimensional chaotic system and a hyperchaotic system are per-
formed to verify the feasibility and effectiveness of the above analytical results.
50
40
30
20
x3
10
0
40
−10 20
−20 0
−10
0 −20
10
20 −40 x
2
x1
Example 4.1. We take the Lorenz system as the first example, which can be described
as follows [11]:
ẋ1 = −ax1 + ax2
ẋ2 = cx1 − x2 − x1 x3 (18)
ẋ3 = −x1 x2 − bx3 ,
To confirm that the complete synchronization is achieved in finite time, we choose the
initial conditions of the Lorenz system as follows: x(0) = [−2, 3, 1]T , y(0) = [3, −2, 2]T ;
and εi (0) = ki (0) = 0, L = 13, k̄ = 3. For simplicity, we take σi (ei (t)) = σei (t), i =
1, 2, . . . , n. Additionally, assume that Ẇ (t) is a one-dimensional white noise. Then,
σi (ei (t)) satisfies the locally Lipschitz condition and the linear growth condition. The
corresponding numerical results are shown in Figures 2 (a) and (b). Figure 2 (a) shows
the temporal evolutions of synchronization errors between Eqs. (1) and (2), and the
temporal evolutions of variable strengths εi , ki are shown in Figure 2 (b). The control
parameters εi and ki converge to -13 and -3 respectively. The simulation results show
that the slave system (2) synchronizes the master system (1) after T1 = 0.7902.
10 0 0
e1
e2
k1
ε1
8 e3 −10 −2
6 −20 −4
0 0.5 1 1.5 2 0 0.5 1 1.5 2
4 0 0
e1, e2, e3
k2
ε2
2 −10 −2
−20 −4
0 0 0.5 1 1.5 2 0 0.5 1 1.5 2
−2 0 0
k3
ε3
−10 −2
−4
−20 −4
−6 0 0.5 1 1.5 2 0 0.5 1 1.5 2
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 t t
t
(a) (b)
To study the effect of the violations of noise intensities σ on the settling time, we sim-
ulate the evolutions of two chaotic systems with the controllers defined in Eq.(7) through
taking different values of σ. Figure 3 gives the evolutions of the total errors function
E(t) with different values of σ, where E(t) = ke(t)k. It shows that the synchronization
time is inversely proportionally to the noise intensity.
Example 4.2. To show the generality of the present method, we take the hyperchaotic
Rössler system as the second example. The hyperchaotic Rössler system can be described
by a four-dimensional differential equation as follows [22, 23]:
ẋ1 = −x2 − x3
ẋ2 = x1 + ax2 + x4
(19)
ẋ3 = x1 x3 + b
ẋ4 = −cx3 + dx4 ,
where x = (x1 , x2 , x3 , x4 )T ∈ R4 is the state vector. System (19) has a chaotic attractor
when a = 0.25, b = 3, c = 0.5, and d = 0.05 as shown in Figure 4.
Finite-time synchronization of chaotic systems 145
12
σ=4
σ=3
σ=2
10
σ=1
8
E(t)
0
0 0.5 1 1.5
t
6000
4000
2000
x3
−6000
0
−4000
−2000
−1
0 −2000
1
2 x2
3 0
x1
For the hyperchaotic Rössler system, we choose the initial conditions of the master
system (1) and the controlled slave system (2) as x(0) = [−4, 1, 3, −1]T and y(0) =
[2, −4, 3, 1]T respectively; and the other parameters as εi (0) = ki (0) = 0, L = 7, k̄ = 2.
At the same time, take σi (ei (t)) = σei (t) (i = 1, 2, . . . , n). From Figures 5 (a) and
(b), one can find that the slave system (2) synchronizes the master system (1) after
T1 = 0.8709. And the control parameters εi and ki converge to -7 and -2 respectively.
8
e1 0 0
e2
k1
ε1
6 e3 −5 −2
e4 −10 −4
0 0.5 1 1.5 2 0 0.5 1 1.5 2
4
0 0
k2
ε2
−5 −2
e1, e2, e3,e4
2
−10 −4
0 0.5 1 1.5 2 0 0.5 1 1.5 2
0 0 0
k3
ε3
−5 −2
−2 −10 −4
0 0.5 1 1.5 2 0 0.5 1 1.5 2
t t
0 0
−4
k4
ε4
−5 −2
−10 −4
−6 0 0.5 1 1.5 2 0 0.5 1 1.5 2
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 t t
t
(a) (b)
The above two examples show that chaotic or hyperchaotic synchronization can be
quickly achieved by the present method (i. e., the settle time is short) with noise pertur-
bation. The time-varying feedback gains εi and ki automatically converge to suitable
constants.
5. CONCLUSIONS
ACKNOWLEDGEMENT
This work was supported by the National Natural Science Foundation of China (grant nos. 61403393)
and the Fundamental Research Funds for the Central Universities (grant no. 2013XK03).
REFERENCES
[1] M. P. Aghababa and H. P. Aghababa: A general nonlinear adaptive control scheme for
finite-time synchronization of chaotic systems with uncertain parameters and nonlinear
inputs. Nonlinear Dyn. 69 (2012), 1903–1914. DOI:10.1007/s11071-012-0395-1
[2] M. P. Aghababa and H. P. Aghababa: A novel finite-time sliding mode controller for
synchronization of chaotic systems with input nonlinearity. Arab. J. Sci. Eng. 38 (2013),
3221–3232. DOI:10.1007/s13369-012-0459-z
[3] M. P. Aghababa, S. Khanmohammadi, and G. Alizadeh: Finite-time synchronization of
two different chaotic systems with unknown parameters via sliding mode technique. Appl.
Math. Model. 35 (2011), 3080–3091. DOI:10.1016/j.apm.2010.12.020
[4] G. Alvarez, L. Hernández, J. Muñoz, F. Montoya, and S. J. Li: Security analysis of
communication system based on the synchronization of different order chaotic systems.
Phys. Lett. A 345 (2005), 245–250. DOI:10.1016/j.physleta.2005.07.083
[5] F. Argenti, A. DeAngeli, E. DelRe, R. Genesio, P. Pagni, and A. Tesi: Secure communi-
cations based on discrete time chaotic systems. Kybernetika 33 (1997), 41–50.
[6] Z. Beran: On characterization of the solution set in case of generalized semiflow. Kyber-
netika 45 (2009), 701–715.
[7] S. Boccaletti, J. Kurths, G. Osipov, D. L. Valladares, and C. S. Zhou: The synchronization
of chaotic systems. Phys. Rep. 366 (2002), 1–101. DOI:10.1016/s0370-1573(02)00137-0
[8] N. Cai, W. Q. Li, and Y. W. Jing: Finite-time generalized synchronization of chaotic
systems with different order. Nonlinear Dyn. 64 (2011), 385–393. DOI:10.1007/s11071-
010-9869-1
[9] S. Cheng, J. C. Ji, and J. Zhou: Fast synchronization of directionally coupled chaotic
systems. Appl. Math. Model. 37 (2013), 127–136. DOI:10.1016/j.apm.2012.02.018
[10] S. Čelikovský: Observer form of the hyperbolic-type generalized Lorenz system and its
use for chaos synchronization. Kybernetika 40 (2004), 649–664.
[11] S. Čelikovský and G. R. Chen: On the generalized Lorenz canonical form. Chaos Solition.
Fract. 26 (2005), 1271–1276. DOI:10.1016/j.chaos.2005.02.040
[12] K. Ding and Q. L. Han: Effects of coupling delays on synchronization in
Lur’e complex dynamical networks. Int. J. Bifur. Chaos 20 (2010), 3565–3584.
DOI:10.1142/s0218127410027908
[13] K. Ding and Q. L. Han: Master-slave synchronization criteria for horizontal platform
systems using time delay feedback control. J. Sound Vibration 330 (2011), 2419–2436.
DOI:10.1016/j.jsv.2010.12.006
[14] K. Ding and Q. L. Han: Master-slave synchronization of nonautonomous chaotic systems
and its application to rotating pendulums. Int. J. Bifur. Chaos 22 (2012), 1250147.
DOI:10.1142/s0218127412501477
148 J. WU, Z. C. MA, Y. Z. SUN AND F. LIU
Jie Wu, School of Sciences, China University of Mining and Technology, Xuzhou,
Jiangsu, 221008. P. R. China.
e-mail: wujiecumt@hotmail.com
Zhi-cai Ma, School of Sciences, China University of Mining and Technology, Xuzhou,
Jiangsu, 221008. P. R. China.
e-mail: zhicai ma@hotmail.com
Feng Liu, School of Sciences, China University of Mining and Technology, Xuzhou,
Jiangsu, 221008. P. R. China.
e-mail: fengliuyjs@163.com
Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.
Alternative Proxies: