You are on page 1of 6

International Journal of Bifurcation and Chaos, Vol. 19, No.

11 (2009) 38233828 c World Scientic Publishing Company

A HYPERCHAOTIC CHUA SYSTEM


PAULO C. RECH and HOLOKX A. ALBUQUERQUE Departamento de F sica, Universidade do Estado de Santa Catarina, 89223-100 Joinville, Brazil d2pcr@joinville.udesc.br d2haa@joinville.udesc.br Received March 24, 2009; Revised April 8, 2009
In this paper, we report a new four-dimensional autonomous hyperchaotic system, constructed from a Chua system where the piecewise-linear function usually taken to describe the nonlinearity of the Chua diode has been replaced by a cubic polynomial. Analytical and numerical procedures are conducted to study the dynamical behavior of the proposed new hyperchaotic system. Keywords : Hyperchaotic Chua system; cubic nonlinearity; phase-space; Lyapunov exponent.

1. Introduction
Hyperchaos was rst presented by R ossler [1979] to characterize a chaotic system with more than one positive Lyapunov exponent. It means that the dynamics of the system is expanded in more than one direction simultaneously, resulting in a more complex chaotic attractor when we compare with the chaotic system with only one positive Lyapunov exponent. This expansion of the dynamics, happening at the same time in two or more directions, leads to hyperchaotic systems showing better performance in many elds, including technological applications, when compared to chaotic systems. For example, hyperchaotic systems, due to higher unpredictability, can be used to improve the security in chaotic communication systems, where a chaotic signal is used to mask the message to be transmitted, once such messages in chaotic systems are not always secure [Perez & Cerdeira, 1995]. As far as we know, no recipe exists to construct a hyperchaotic system. An ecient method considers at least a fourth dimension in a threedimensional chaotic system, by the introduction of

feedback controllers. It was done recently with a Lorenz system [Barboza, 2007; Cai et al., 2008], a Chen system [Gao et al., 2006; Gao et al., 2009; Li et al., 2005], and L u system [Bao & Liu, 2008; Chen et al., 2006; Wang et al., 2006]. In this work, a new hyperchaotic system is constructed, from a Chua system where the piecewise-linear function usually taken to describe the nonlinearity of the Chua diode has been replaced by a cubic polynomial. The paper is organized as follows. In Sec. 2, we present the new system, obtained from a Chua system with a cubic nonlinearity by the inclusion of a fourth variable. In Sec. 3, we discuss some basic properties of system (2), and show that the trivial xed point can be unstable. In Sec. 4, numerical results involving Lyapunov exponents and phasespace portraits are presented. The results are summarized in Sec. 5.

2. The New System


The set of equations of motion that describes theoretically the Chua oscillator with a smooth nonlinearity dened by g(x) = ax3 + cx, can be easily

3823

3824

P. C. Rech & H. A. Albuquerque

obtained in a dimensionless form [Albuquerque & Rech, 2009; Bilotta et al., 2007] and is given by x = (y ax3 (1 + c)x), y = x y + z, z = y z, (1)

where x, y , z represent dynamical variables, and , , , a, c are control parameters. The constants , , are dierent functions of the circuit linear components [Huang et al., 1996], while a and c dene the g(x) function. Based on Chua oscillator, we construct a new four-dimensional system by introducing a fourth variable w, which is an adequate feedback controller to the third equation in system (1), to obtain x = (y ax3 (1 + c)x), y = x y + z, z = y z + w, w = sx + yz,

for x, y , z and w. Obviously, the origin P0 (0, 0, 0, 0) is a xed point. Not so obvious is the fact that system (3) does not present an analytical solution for the other xed points. Hence, in sequence, we analyze the stability of the origin P0 . The Jacobian matrix for system (2) at P0 , denoted by J0 , is given by (1 + c) 0 0 1 1 1 0 J0 = , 0 1 s 0 0 0 and the eigenvalue equation, calculated using det(J0 mI ) = 0, where I is the 4 4 identity matrix and m represents the eigenvalues, is Am4 + Bm3 + Cm2 + Dm + E = 0, (4)

(2)

where A = 1, B = (1 + c) + + 1, C = [ + c(1 + )] + + , D = [ + c( + )], E = s.

where s is another parameter. As we will see in Sec. 4, when the parameters are properly chosen, system (2) can exhibit hyperchaos.

3. Analytical Results
The divergence [Ott, 2000] of the four-dimensional system (2) is given by y z w x + + + = (3ax2 + c + 1) 1, x y z w and from this result we conclude that there are combinations of parameters for which the system is dissipative, that is, for which (3ax2 + c+1)+ +1 > 0. It means that the phase-space contracts volumes at an exponential rate equal to e[(3ax
2 +c+1)+ +1]t

Some information on the roots of polynomial (4) and, as a consequence about the eigenvalues associated with the origin P0 , can be obtained just by looking at the coecients A, B , C , D and E . Parameters , a, c, , and s can be appropriately chosen such that sign changes from one coecient to the next take place in the sequence E A. Descartes Rule of Signs [Wiggins, 2003] states that the number of positive real roots of the polynomial

z
8 4 0 -4 -8 0.6 -4 0.2 -2 -0.2 4 -0.6

as time t increases. As a consequence, all the system trajectories nally converge to an attractor in a four-dimensional phase-space. Therefore, we conclude that attractors are possible in system (2). The xed points are obtained when x = 0, y = 0, z = 0, and w = 0 in Eqs. (2), that is, by solving the set of algebraic equations (y ax3 (1 + c)x) = 0, x y + z = 0, y z + w = 0, sx + yz = 0,

x
(3)

Fig. 1. Chaotic attractor in three-dimensional phase-space of the system (1) with = 0.32, a = 0.03, c = 1.2, = 30, = 50, from the initial condition (x, y, z ) = (0.0, 0.1, 0.3), constructed with 2 104 points.

A Hyperchaotic Chua System

3825

(4) is equal to the total number of sign changes. Thus, according to Descartes Rule, not all eigenvalues have negative real parts. As is well known, the origin P0 is a stable xed point if the real part of the corresponding eigenvalues is negative. Our conclusion is that the origin P0 can be unstable depending on the parameters and, therefore, chaos (or hyperchaos) is possible in system (2).

4. Numerical Results
When = 0.32, a = 0.03, c = 1.2, = 30, = 50, Chua oscillator with a cubic nonlinearity as dened by Eqs. (1) has a typical chaotic attractor as seen in Fig. 1. The corresponding Lyapunov exponents are 1 = 0.7473, 2 = 0.0000, 3 = 17.2204, and the Lyapunov dimension (also called KaplanYorke dimension) is DL = 2.0434. The system (1) was

0.8

2
0.0

1
-0.2

0.4

-0.4 0.0 0.1


(a)

0.2

0.3

-0.6

0.1

s
(b)

0.2

0.3

-50

-100

0.1
(c)

0.2

0.3

Fig. 2. Lyapunov exponents spectrum of system (2), as a function of s with 0.1 s 0.3, = 0.32, a = 0.03, c = 1.2, = 30, = 50. The double arrow in (b) indicates the hyperchaotic region, and the dashed line in black in (b) and (c) locates = 0.

3826

P. C. Rech & H. A. Albuquerque

integrated with a fourth-order RungeKutta algorithm with a xed step size equal to 0.01, and considering 5 105 steps to compute the Lyapunov exponents. The spectrum of Lyapunov exponents for system (2), integrated with the same conditions as system (1), as a function of the parameter s, is shown in Fig. 2. It can be observed in Figs. 2(a) and 2(b) that the parameter s covers a small region, limited by the double arrow in Fig. 2(b), from s 0.1060 to s 0.1631, for which the system (2) has two positive Lyapunov exponents (1 and 2 ), that is, the system (2) is hyperchaotic. For other values of 0.1 s 0.3, there is only one positive Lyapunov exponent (1 ), and the system is chaotic. Thus, for s < 0.1060 only the largest Lyapunov exponent 1 is positive, and the system (2) evolves on a chaotic attractor. When the parameter s is increased so that the second largest Lyapunov exponent 2 becomes positive in

s 0.1060, the attractor becomes hyperchaotic. After passing the short hyperchaotic window, the system (2) evolves into a chaotic motion again in s 0.1631, when 2 touches zero. The transition chaoshyperchaos is a consequence of changes in the stability of an innite number of unstable periodic orbits embedded in the chaotic attractor, which become repellers in the neighborhood of the transition [Kapitaniak et al., 2000]. In Figs. 2(b) and 2(c) we can see the behavior of the third largest Lyapunov exponent, 3 , and the smaller Lyapunov exponent, 4 . Three-dimensional projections of the fourdimensional attractor for the new hyperchaotic system (2) are shown in Figs. 3(a)3(d), for s = 0.1111. The corresponding Lyapunov exponents are 1 = 0.7325, 2 = 0.0145, 3 = 0.0000, 4 = 17.7662, and the Lyapunov dimension is DL = 3.0420. In Fig. 3(a), we can see that the projection of the hyperchaotic trajectory on three-dimensional

z
8 4 0 -4 -8 0.6 -4 0.2 -2 -0.2 2
(a)

w
2 1 0 -1 -2 -3 0.6 -4 0.2 -2

-0.6

-0.2 2
(b)

-0.6

w
2 1 0 -1 -2 -3 -4 -2 0 4 8
8 4 0 -4 -8

0.6 -3 0.2 -2

2
(c)

4 -8

-4

-1

-0.2 0 1
(d)

-0.6

Fig. 3. Three-dimensional projections of the hyperchaotic attractor generated by system (2), from the initial condition (x, y, z, w) = (0.0, 0.1, 0.3, 0.4). All diagrams were constructed with 2 104 points. (a) xyz phase space, in red. (b) xyw phase space, in green. (c) xzw phase space, in blue. (d) yzw phase space, in black.

A Hyperchaotic Chua System

3827

xyz phase-space is double-scroll-like, which is typical of Chua system (1). The only threedimensional projection where this double-scroll-like structure does not appear is the yzw projection in Fig. 3(d). In Fig. 4, one sees all the two-dimensional projections of the new hyperchaotic attractor (2), also

for = 0.32, a = 0.03, c = 1.2, = 30, = 50, s = 0.1111. The objective to include these portraits here, is that they provide another opportunity to observe the structure of the orbit more clearly. Again the double-scroll-like structure of the plane projections is clearly visible in Fig. 4(a) (xy plane), Fig. 4(b) (xz plane), and Fig. 4(d) (yz plane).

(a)

(b)

(c)

(d)

Fig. 4. Two-dimensional projections of the hyperchaotic attractor generated by system (2), from the initial condition (x, y, z, w) = (0.0, 0.1, 0.3, 0.4). All diagrams were constructed with 2 104 points. (a) xy phase plane, in black. (b) xz phase plane, in blue. (c) xw phase plane, in violet. (d) yz phase plane, in brown. (e) yw phase plane, in magenta. (f) zw phase plane, in red.

3828

P. C. Rech & H. A. Albuquerque

(e) Fig. 4. (Continued )

(f)

5. Summary
In summary, we report a new four-dimensional autonomous hyperchaotic system, coined from a Chua system where the piecewise-linear function usually taken to describe the nonlinearity of the Chua diode has been replaced by a cubic polynomial. Analytical investigations concerning dissipation, xed points and stability of the trivial xed point have been realized. Numerical procedures involving Lyapunov exponents and phase-space portraits were utilized to characterize this new hyperchaotic attractor.

Acknowledgments
The authors thank Conselho Nacional de Desenvolvimento Cient co e Tecnol ogico-CNPq, and Funda c ao de Apoio a ` Pesquisa Cient ca e Tecnol ogica-FAPESC, Brazilian agencies, for the nancial support.

References
Albuquerque, H. A. & Rech, P. C. [2009] A parameterspace of a Chua system with a smooth nonlinearity, Int. J. Bifurcation and Chaos 19, 13511355. Bao, B.-C. & Liu, Z. [2008] A hyperchaotic attractor coined from chaotic L u system, Chin. Phys. Lett. 25, 23962399. Barboza, R. [2007] Dynamics of a hyperchaotic Lorenz system, Int. J. Bifurcation and Chaos 17, 4285 4294.

Bilotta, E., Di Blasi, G., Stranges, F. & Pantano, P. [2007] A gallery of Chua attractors. Part IV, Int. J. Bifurcation and Chaos 17, 10171077. Cai, G.-L., Zheng, S. & Tian, L.-X. [2008] Adaptive control and synchronization of an uncertain new hyperchaotic Lorenz system, Chin. Phys. B 17, 24122419. Chen, A., Lu, J., L u, J. L. & Yu, S. [2006] Generating hyperchaotic L u attractor via state feedback control, Physica A 364, 103110. Gao, T. G., Chen, Z. Q. & Chen, G. [2006] A hyperchaos generated from Chens system, Int. J. Mod. Phys. C 17, 471478. Gao, T., Gu, Q. & Chen, Z. [2009] Analysis of the hyper-chaos generated from Chens system, Chaos Solit. Fract. 39, 18491855. Huang, A., Pivka, L., Wu, C. W. & Franz, M. [1996] Chuas equation with cubic nonlinearity, Int. J. Bifurcation and Chaos 6, 21752222. Kapitaniak, T., Maistrenko, Y. & Popovych, S. [2000] Chaos-hyperchaos transition, Phys. Rev. E 62, 19721976. Li, Y., Tang, S. K. & Chen, G. [2005] Generating hyperchaos via state feedback control, Int. J. Bifurcation and Chaos 15, 33673375. Ott, E. [2000] Chaos in Dynamical Systems (Cambridge University Press). Perez, G. & Cerdeira, H. A. [1995] Extracting messages masked by chaos, Phys. Rev. Lett. 74, 19701973. R ossler, O. E. [1979] An equation for hyperchaos, Phys. Lett. A 71, 155157. Wang, G., Zhang, X., Zheng, Y. & Li, Y. [2006] A new modied hyperchaotic L u system, Physica A 371, 260272. Wiggins, S. [2003] Introduction to Applied Nonlinear Dynamical Systems and Chaos (Springer).