1: %Version of 22 April 2003
2: \documentclass[twocolumn,showpacs,preprintnumbers,amsmath,amssymb]{revtex4}
3:
4: \usepackage{graphicx}
5: \usepackage{dcolumn}
6: \usepackage{bm}
7: \graphicspath{{./}{./figs/}}
8:
9:
10: \begin{document}
11:
12: \title{\bf Synchronization of fractional order chaotic systems \footnote{Phys. Rev. E, Vol. 68, 067203, 2003}}
13: \author{Chunguang Li$^1$, Xiaofeng Liao$^{1,2}$, Juebang Yu$^1$}
14: %\email{cgli@uestc.edu.cn}
15: \affiliation{$^1$Institute of Electronic Systems, School of
16: Electronic Engineering,\\ University of Electronic Science and
17: Technology of China,\\ Chengdu, Sichuan, 610054, P. R. China.\\
18: $^2$School of Computer Science and Technology, Chongqing University,\\
19: Chongqing, 400044, P. R. China.}
20: \date{}
21: \begin{abstract}
22: The chaotic dynamics of fractional order systems begin to attract
23: much attentions in recent years. In this brief report, we study
24: the master-slave synchronization of fractional order chaotic
25: systems. It is shown that fractional order chaotic systems can
26: also be synchronized.
27: \end{abstract}
28: \pacs{05.45.-a}
29: \maketitle
30:
31: Fractional calculus is a 300-year-old topic. Although it has a
32: long mathematical history, the applications of fractional calculus
33: to physics and engineering are just a recent focus of interest [1,
34: 2]. Many systems are known to display fractional order dynamics,
35: such as viscoelastic systems [3-5], dielectric polarization [6],
36: electrode-electrolyte polarization [7] and electromagnetic waves
37: [8]. More recently, many authors begin to investigate the chaotic
38: dynamics of fractional order dynamical systems [9-17]. In [9], it
39: has been shown that the fractional order Chua's system of order as
40: low as 2.7 can produce a chaotic attractor. In [10], it has been
41: shown that nonautonomous Duffing systems of order less than 2 can
42: still behave in a chaotic manner. In [11], chaotic behaviors of
43: the fractional order ``jerk" model was studied, in which chaotic
44: attractor was obtained with system orders as low as 2.1, and in
45: [12] the chaos control of this fractional order chaotic system was
46: reported. In [13], chaotic behavior of the fractional order Lorenz
47: system was studied, but unfortunately, the results presented in
48: this paper are not correct. In [14] and [15], bifurcation and
49: chaotic dynamics of the fractional order cellular neural networks
50: were studied. In [16], chaos and hyperchaos in the fractional
51: order R\"ossler equations were studied, in which we showed that
52: chaos can exists in the fractional order R\"ossler equation with
53: order as low as 2.4, and hyperchaos exists in the fractional order
54: R\"ossler hyperchaos equation with order as low as 3.8. In [17],
55: we have studied the chaotic behavior and its control in the
56: fractional order Chen system. In [18], the author present a broad
57: review of existing models of fractional kinetics and their
58: connection to dynamical models, phase space topology, and other
59: characteristics of chaos.
60:
61: On the other hand, synchronization of chaotic systems has
62: attracted much attentions [19] since the seminal paper by Pecora
63: and Carroll [20]. In this brief report, we study the
64: synchronization of fractional order chaotic systems. The analysis
65: of fractional order systems is by no means trivial. So, we will
66: numerically investigate this topic here.
67:
68: There are many definitions of fractional derivatives [1]. Perhaps
69: the best known one is the Riemann-Liouville definition, which is
70: given by
71: \begin{equation}
72: \frac{d^\alpha f(t)}{dt^\alpha} = \frac{1} {\Gamma(n-\alpha)}
73: \frac{d^n}{dt^n} \int_0^t
74: \frac{f(\tau)}{(t-\tau)^{\alpha-n+1}}d\tau
75: \end{equation}
76: where $\Gamma(\cdot)$ is the gamma function and $n-1\leq \alpha
77: <n$. The geometric and physical interpretation of the fractional
78: derivatives was given in [21]. Upon considering the initial
79: conditions to be zero, the Laplace transform of the
80: Riemann-Liouville fractional derivative is $
81: L\left\{\frac{d^\alpha f(t)}{dt^\alpha}\right\}=s^\alpha
82: L\{f(t)\}$. So, the fractional integral operator of order
83: ``$\alpha$" can be represented by the transfer function
84: $F(s)=\frac{1}{s^\alpha}$.
85:
86: The standard definition of the fractional differintegral do not
87: allow direct implementation of the fractional operators in
88: time-domain simulations. An efficient method to circumvent this
89: problem is to approximate the fractional operators by using the
90: standard integer order operators. In the following simulations, we
91: will use the approximation method proposed in [22], which was also
92: adopted in [9, 11, 14, 15, 16, 17]. In Table 1 of [9], the authors
93: gave approximations for $1/s^q$ with $q=0.1-0.9$ in steps 0.1 with
94: errors of approximately 2dB. We will use these approximations in
95: our following simulations.
96:
97: Consider the master-slave synchronization scheme of two autonomous
98: $n$-dimensional fractional order chaotic systems
99: \begin{equation}
100: \begin{array}{rl}
101: M&: \frac{d^\alpha x}{dt^\alpha}=f(x)\\
102: S&: \frac{d^\alpha y}{dt^\alpha}=f(y)+c\Gamma(x-y)
103: \end{array}
104: \end{equation}
105: with the master system $M$ and the slave system $S$. Where
106: $\alpha>0$ is the fractional order, with which the individual
107: dynamical systems are chaotic, $c>0$ is the coupling strength, and
108: $\Gamma\in R^{n\times n}$ is a constant 0 - 1 matrix linking the
109: coupling variables. For simplicity, we assume
110: $\Gamma=\mbox{diag}(r_1,r_2,\cdots,r_n)$ is a diagonal matrix. If
111: there is a coupling between the $i$th state variable of the two
112: coupled chaotic systems, then $r_i=1$; otherwise, $r_i=0$. Define
113: the error signal as $e=x-y$, the aim of the synchronization scheme
114: is to design the coupling strength such that $\|e(t)\|\rightarrow
115: 0$ as $t\rightarrow \infty$. This scheme is similar to the
116: master-slave synchronization of classical integer-order chaotic
117: systems.
118:
119: Next, we numerically study the synchronization of fractional order
120: chaotic systems via two examples. we first consider the fractional
121: order Chua's system [9]
122: \begin{equation}
123: \begin{array}{l}
124: \frac{d^\alpha x}{dt^\alpha} =a\left[y+\frac{x-2x^3}{7}\right]\\
125: \frac{d^\alpha y}{dt^\alpha}=x-y+z\\
126: \frac{d^\alpha z}{dt^\alpha}=-\frac{100}{7}y
127: \end{array}
128: \end{equation}
129: when $\alpha\geq 0.9$, this system can produce chaotic solutions
130: [9]. Particularly, when $\alpha=0.9$ and $a=12.75$, the fractional
131: order Chua's system is chaotic. The phase plot of $x$ and $z$ is
132: shown in Fig.1.
133: \begin{figure}[htb]
134: \centering
135: \includegraphics[width=6cm]{fig1.eps}
136: \caption{Phase plot of the fractional order Chua's system with
137: $\alpha=0.9$.}
138: \end{figure}
139: \begin{figure}[htb]
140: \centering
141: \includegraphics[width=8.5cm]{fig2.eps}
142: \caption{Synchronization error of the fractional order Chua's
143: systems with $\alpha=0.9$: (a) $c=4$, (b) $c=7$.}
144: \end{figure}
145: \begin{figure}[htb]
146: \centering
147: \includegraphics[width=8.5cm]{fig3.eps}
148: \caption{Synchronization error of the integer order Chua's
149: systems: (a) $c=4$, (b) $c=7$.}
150: \end{figure}
151:
152:
153: We let $\Gamma=\mbox{diag}(1,0,0)$, which implies that only the
154: first variable $x$ is used to coupling the two fractional order
155: chaotic systems. To obtain a critical value of $c$ to make the two
156: systems synchronized, we continuously increase the coupling
157: strength $c$, from $c=0$, in step 0.5. When $c<4$, no synchronous
158: phenomenon is observed. When $c=4$, the curve of the
159: synchronization error $J(t)=\mbox{log}(\|e(t)\|)$ is shown in
160: Fig.2 (a), which indicate that the master-slave synchronization is
161: achieved. In Fig.2 (b), we show the curve of the synchronization
162: error when $c=7$, in which the synchronization effect is better
163: than that of $c=4$.
164:
165: For the purpose of comparison, we also plot the curves of
166: synchronization error of the integer order Chua's systems
167: ($a=9.5$) in Fig.3. Comparing Fig. 2 with Fig.3, we can know that
168: the synchronization rate of the fractional order Chua's systems is
169: slower than its integer order counterpart.
170:
171: We next consider the fractional order R\"ossler system [16]:
172: \begin{equation}
173: \begin{array}{l}
174: \frac{d^\alpha x}{dt^\alpha} =-(y+z)\\
175: \frac{d^\alpha y}{dt^\alpha}=x+ay\\
176: \frac{d^\alpha z}{dt^\alpha}=0.2+z(x-10)
177: \end{array}
178: \end{equation}
179: when $\alpha=0.9$ and $a=0.4$, the above system is chaotic. The
180: phase diagram of the chaotic attractor is shown in Fig. 4.
181:
182: We also let $\Gamma=\mbox{diag}(1,0,0)$, and do the similar
183: simulations as in the above example. When the coupling strength
184: $c=0.5$, the two fractional R\"ossler systems achieve
185: synchronization. The curve of the synchronization error of the
186: fractional order R\"ossler system is shown in Fig. 5 (a). In Fig.
187: 5 (b), we also plot the curve of the synchronization error of the
188: integer order R\"ossler systems ($a=0.165$). From Fig. 5, we know
189: that the synchronization rate of the fractional order R\"ossler
190: systems is also slightly slower than its integer counterpart.
191: \begin{figure}[htb]
192: \centering
193: \includegraphics[width=7cm]{fig4.eps}
194: \caption{Phase plot of the fractional order R\"ossler system with
195: $\alpha=0.9$.}
196: \end{figure}
197: \begin{figure}[htb]
198: \centering
199: \includegraphics[width=8.5cm]{fig5.eps}
200: \caption{The curves of synchronization error: (a) the fractional
201: order R\"ossler systems with $\alpha=0.9$ and $c=0.5$; (b) the
202: integer order R\"ossler systems with $c=0.5$.}
203: \end{figure}
204:
205: We have also tested the synchronization scheme (2) on several
206: other fractional order chaotic systems [23]. Limited to the length
207: of this brief report, we omit these results here.
208:
209: In summary, in this brief report, we have studied the master-slave
210: synchronization of fractional order chaotic systems. To our best
211: knowledge, this is the first report on the synchronization of
212: fractional order dynamical systems. We have shown that fractional
213: order chaotic systems can be synchronized by utilizing the similar
214: scheme as that of their integer order counterparts.
215:
216: Future works regarding this topic include the investigation of
217: some other types of synchronization of fractional order chaotic
218: systems, such as the phase synchronization [24] and the projective
219: synchronization [25], as well as the synchronization of fractional
220: order hyperchaotic systems [16].
221:
222: We acknowledge supports from the National Natural Science
223: Foundation of China under Grant 60271019, and the Youth Science
224: and Technology Foundation of UESTC under Grant YF020207.
225:
226: \begin{thebibliography}{20}
227: \bibitem{"[1]"}
228: I. Podlubny, {\it Fractional Differential Equations}, (Academic
229: Press, New York, 1999).
230: \bibitem{"[2]"}
231: R. Hilfer (Ed.), {\it Applications of Fractional Calculus in
232: Physics}, (World Scientific, New Jersey, 2001).
233: \bibitem{"[3]"}
234: R.L. Bagley, and R.A. Calico, J. Guid., Contr. Dyn. 14, 304
235: (1991).
236: \bibitem{"[4]"}
237: R. C. Koeller, J. Appl. Mech. 51, 299 (1984).
238: \bibitem{"[5]"}
239: R.C. Koeller, Acta Mechanica 58, 251 (1986).
240: \bibitem{"[6]"}
241: H.H. Sun, A.A. Abdelwahad, and B. Onaral, IEEE Trans. Auto. Contr.
242: 29, 441 (1984).
243: \bibitem{"[7]"}
244: M. Ichise, Y. Nagayanagi, and T. Kojima, J. Electroanal. Chem. 33,
245: 253 (1971).
246: \bibitem{"[8]"}
247: O. Heaviside, {\it Electromagnetic Theory}, (Chelsea, New York,
248: 1971).
249: \bibitem{"[9]"}
250: T.T. Hartley, C.F. Lorenzo, and H. K. Qammer, IEEE Trans. CAS-I
251: 42, 485 (1995).
252: \bibitem{"[10]"}
253: P. Arena, R. Caponetto, L. Fortuna, and D. Porto, Proc. ECCTD,
254: Budapest, 1259 (1997).
255: \bibitem{"[11]"}
256: W.M. Ahmad, and J.C. Sprott, Chaos, Solitons and Fractals 16, 339
257: (2003).
258: \bibitem{"[12]"}
259: W.M. Ahmad, and W.M. Harb, Chaos, Solitons and Fractals 18, 693
260: (2003).
261: \bibitem{"[13]"}
262: I. Grigorenko, and E. Grigorenko, Phys. Rev. Lett. 91, 034101
263: (2003). But the Eq. (5) in this paper is not correct, so the
264: results presented in this paper are not reliable, which we have
265: pointed out to the first author of this paper via personal
266: communication.
267: \bibitem{"[14]"}
268: P. Arena, R. Caponetto, L. Fortuna, and D. Porto, Int. J. Bifur.
269: Chaos 7, 1527 (1998).
270: \bibitem{"[15]"}
271: P. Arena, L. Fortuna, and D. Porto, Phys. Rev. E 61, 776 (2000).
272: \bibitem{"[16]"}
273: C. Li, G. Chen, X. Liao, and J. Yu, Submitted (2003).
274: \bibitem{"[17]"}
275: C. Li, G. Chen, Submitted (2003).
276: \bibitem{"[18]"}
277: G.M. Zaslavsky, Phys. Rep. 371, 461 (2002).
278: \bibitem{"[19]"}
279: G. Chen, {\it Control and Synchronization of Chaos, a
280: Bibliography}, (ECE Dept, Univ of Houston, TX -- available from
281: ftp: ``ftp.egr.uh.edu/pub/Tex/chaos.tex'' (login name
282: ``anonymous'' password: your email address), 1997).
283: \bibitem{"[20]"}
284: L.M. Pecora and T.L. Carroll, Phys. Rev. Lett. 64, 821 (1990).
285: \bibitem{"[21]"}
286: I. Podlubny, arXiv: math.CA/0110241 v1 (2001).
287: \bibitem{"[22]"}
288: A. Charef, H.H. Sun, Y.Y. Tsao, and B.
289: Onaral, IEEE Trans. Auto. Contr. 37, 1465 (1992).
290: \bibitem{"[23]"}
291: ``Several other fractional order chaotic systems" in this paper
292: include the fractional order ``jerk" model [11], the fractional
293: order Chen system [17], and the fractional order CNN [14, 15].
294: \bibitem{"[24]"}
295: M. Rosenblum, A. Pikovsky, and J. Kurtz, Phys. Rev. Lett. 76, 1804
296: (1996); A. Pikovsky, M. Rosenlum, G. Osipov, and J. Kurtz, Phys. D
297: 104, 219 (1997); S. Boccaletti, J. Kurths, G. Osipov, D.L.
298: Valladares, and C.S. Zhou, Phys. Rep. 366, 1 (2002).
299: \bibitem{"[25]"}
300: R. Mainieri and J. Rehacek, Phys. Rev. Lett. 82, 3042 (1999).
301: \end{thebibliography}
302: \end{document}
303: