- •LIBRARY OF CONGRESS CATALOGING-IN-PUBLICATION DATA
- •CONTENTS
- •FOREWORD
- •PREFACE
- •Abstract
- •1. Introduction
- •2. A Nice Equation for an Heuristic Power
- •3. SWOT Method, Non Integer Diff-Integral and Co-Dimension
- •4. The Generalization of the Exponential Concept
- •5. Diffusion Under Field
- •6. Riemann Zeta Function and Non-Integer Differentiation
- •7. Auto Organization and Emergence
- •Conclusion
- •Acknowledgment
- •References
- •Abstract
- •1. Introduction
- •2. Preliminaries
- •3. The Model
- •4. Numerical Simulations
- •5. Synchronization
- •6. Conclusion
- •Acknowledgments
- •References
- •Abstract
- •1. Introduction: A Short Literature Review
- •2. The Injection System
- •3. The Control Strategy: Switching of Fractional Order Controllers by Gain Scheduling
- •4. Fractional Order Control Design
- •5. Simulation Results
- •6. Conclusion
- •Acknowledgment
- •References
- •Abstract
- •Introduction
- •1. Basic Definitions and Preliminaries
- •Conclusion
- •Acknowledgments
- •References
- •Abstract
- •1. Context and Problematic
- •2. Parameters and Definitions
- •3. Semi-Infinite Plane
- •4. Responses in the Semi-Infinite Plane
- •5. Finite Plane
- •6. Responses in Finite Plane
- •7. Simulink Responses
- •Conclusion
- •References
- •Abstract
- •1. Introduction
- •2. Modelling
- •3. Temperature Control
- •4. Conclusion
- •References
- •Abstract
- •1. Introduction
- •2. Preliminaries
- •3. Second Order Sliding Mode Control Strategy
- •4. Adaptation Law Synthesis
- •5. Numerical Studies
- •Conclusion
- •References
- •Abstract
- •1. Introduction
- •2. Rabotnov’s Fractional Operators and Main Formulas of Algebra of Fractional Operators
- •4. Calculation of the Main Viscoelastic Operators
- •5. Relationship of Rabotnov Fractional Operators with Other Fractional Operators
- •8. Application of Rabotnov’s Operators in Problems of Impact Response of Thin Structures
- •9. Conclusion
- •Acknowledgments
- •References
- •Abstract
- •1. Introduction
- •3. Theory of Diffusive Stresses
- •4. Diffusive Stresses
- •5. Conclusion
- •References
- •Abstract
- •Introduction
- •Methods
- •Conclusion
- •Acknowledgment
- •Abstract
- •1. Introduction
- •2. Basics of Fractional PID Controllers
- •3. Tuning Methodology for Fuzzy Fractional PID Controllers
- •4. Optimal Fuzzy Fractional PID Controllers
- •5. Conclusion
- •References
- •INDEX
In: Fractional Calculus: Applications |
ISBN: 978-1-63463-221-8 |
Editors: Roy Abi Zeid Daou and Xavier Moreau |
© 2015 Nova Science Publishers, Inc. |
Chapter 2
DYNAMICS OF FRACTIONAL ORDER
CHAOTIC SYSTEMS
Sachin Bhalekar
Department of Mathematics, Shivaji University, Kolhapur, India
Abstract
This chapter deals with the fractional order generalization of the chaotic system proposed by Li et al [Nonlinear Analysis: Modelling & Control, 18 (2013) 66-77]. We discuss the dynamical properties such as symmetry, dissipativity, stability of equilibrium points and chaos. Further we control the chaos in proposed system and present the synchronization phenomenon.
Keywords: Chaos, fractional order, stability
AMS Subject Classification: 26A33, 65P20, 34L30
1.Introduction
Differential equation is one of the basic tools to model the motion of a physical system. When the solutions of these differential equations are bounded, they either settle down to an equilibrium state or oscillates in a periodic or quasi-periodic state. If the order of the system is three or more and the system is nonlinear then there is a possibility of another type of solutions. Such solutions exhibit aperiodic motion for all time and never settles. These solutions are highly sensitive to initial conditions also. Thus, it is not possible to predict the behavior of the solution for a long time. Such systems are termed as chaotic dynamical systems [1].
If the order of the derivative in the system is not an integer then it is called as fractional order system [2, 3, 4]. There are two very important advantages of the fractional order derivative over a conventional integer order derivative. The first advantage is the nonlocal nature of fractional order derivative. The integer order derivative of a function at a point can be approximated using the nearby values whereas the fractional derivative needs the
E-mail addresses: sbb maths@unishivaji.ac.in, sachin.math@yahoo.co.in
Complimentary Contributor Copy
24 |
Sachin Bhalekar |
total history from an initial point. This nonlocality of fractional derivative (FD) is playing a vital role in modelling memory and hereditary properties in the system. Hence the models involving FD are more realistic than those involving integer order ones. Second advantage is the FD is capable of modelling intermediate processes. In some physical problems such as the diffusion of fluid in porous media, the integer order derivative cannot model the real phenomena because the processes are intermediate. In such case, one has to consider a model involving the intermediate (fractional) order derivative.
The generalized dynamical systems involving fractional order derivative are extensively studied by researchers. In the seminal paper, Grigorenko and Grigorenko [5] proposed a fractional order generalization of Lorenz system. Further, various fractional order systems such as Chen system [6], Lu¨ system [7], Rossler system [8]and Liu [9] system are studied in the literature.
In this chapter, we propose a fractional order generalization of the chaotic system proposed by Li et al. [17] and discuss its properties.
2.Preliminaries
2.1.Stability Analysis
We discuss the stability results of the following fractional order system
Dα1 x1 |
= f1(x1, x2, · · · , xn), |
|
Dα2 x2 |
= f2(x1, x2, · · · , xn), |
|
|
. |
|
|
. |
|
|
. |
|
DαN xn |
= fn(x1, x2, · · · , xn), |
(2.1) |
where 0 < αi < 1 are fractional orders. The system (2.1) is called as a commensurate order if α1 = α2 = · · · = αn otherwise an incommensurate order.
A point p = (x1, x2, · · · , xn) is called an equilibrium point of system (2.1) if fi(p) = 0 for each i = 1, 2, · · · , n.
Theorem 2.1. [10, 11] Consider α = α1 = α2 = · · · = αn in (2.1). An equilibrium point p of the system (2.1) is locally asymptotically stable if all the eigenvalues of the Jacobian matrix
|
∂1f1 |
(p) ∂2f1 |
(p) |
· · · |
∂nf1 |
||||||||||
|
|
∂ f |
(p) ∂ f |
(p) |
· · · |
∂nf |
|||||||||
J = |
|
|
1 |
|
2 |
|
2 |
|
2 |
|
|
|
|
2 |
|
|
|
|
.. |
|
|
|
.. |
|
.. |
|
|
|
.. |
||
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
∂ |
1 |
f |
n |
(p) ∂ |
2 |
f |
n |
(p) |
· · · |
∂ |
n |
f |
n |
|
|
|
|
|
|
|
|
|
evaluated at p satisfy the following condition
|arg(Eig(J))| > απ/2.
(p) (p)
(p)
.
(2.2)
(2.3)
Theorem 2.2. [12, 13] Consider the incommensurate fractional ordered dynamical system given by (2.1). Let αi = vi/ui, (ui, vi) = 1, ui, vi be positive integers. Define M to be the least common multiple of ui's.
Complimentary Contributor Copy
Dynamics of Fractional Order Chaotic Systems |
25 |
|
Define |
λMα1 , λMα2 , · · · , λMαN − J |
|
Δ(λ) = diag |
(2.4) |
where J is the Jacobian matrix as defined in (2.2) evaluated at point p. Then p is locally asymptotically stable if all the roots of the equation det (Δ(λ)) = 0 satisfy the condition
|arg(λ)| > π/(2M).
This condition is equivalent to the following inequality |
|
|
π |
− mini {|arg(λi)|} < 0. |
(2.5) |
2M |
Thus an equilibrium point p of the system (2.1) is asymptotically stable if the condition
(2.5) is satisfied. The term |
π |
− mini {|arg(λi)|} is called as the instability measure for |
2M |
equilibrium points in fractional order systems (IMFOS). Hence, a necessary condition for fractional order system (2.1) to exhibit chaotic attractor is [12]
IMFOS ≥ 0. |
(2.6) |
Theorem 2.3. [18] Consider the polynomial |
|
P (λ) = λ3 + a1λ2 + a2λ + a3 = 0. |
(2.7) |
Define the descriminant for equation (2.7) as |
|
D(P ) = 18a1a2a3 + (a1a2)2 − 4a3(a1)3 − 4(a2)3 − 27(a3)2. |
(2.8) |
1. If D(P ) > 0 then all the roots of P (λ) satisfy the condition |
|
|arg(λ)| > απ/2 |
(2.9) |
where 0 ≤ α ≤ 1. |
|
2.If D(P ) < 0, a1 ≥ 0, a2 ≥ 0, a3 > 0, α < 2/3 then (2.9) is satisfied.
3.If D(P ) < 0, a1 > 0, a2 > 0, a1a2 = a3 then (2.9) is satisfied for all 0 ≤ α < 1.
2.2.Numerical Method for Solving Fractional Differential Equations
The numerical methods developed for solving equations involving integer order derivatives are not useful to solve those involving fractional derivatives. A modification of Adams- Bashforth-Moulton algorithm is proposed by Diethelm et al. in [14, 15, 16] to solve FDEs.
Consider for α (m − 1, m] the initial value problem (IVP)
Dαy(t) |
= f (t, y(t)) , 0 ≤ t ≤ T, |
(2.10) |
y(k)(0) |
= y0(k), k = 0, 1, · · · , m − 1. |
(2.11) |
The IVP (2.10)–(2.11) is equivalent to the Volterra integral equation
|
m−1 |
(k) tk |
1 |
t |
|
|
|
|
||
y(t) = |
|
(t − τ)α |
− |
1 f (τ, y(τ)) dτ. |
(2.12) |
|||||
|
X |
|
|
|
|
|
|
|
|
|
k=0
Complimentary Contributor Copy
26 Sachin Bhalekar
Consider the uniform grid {tn = nh/n = 0, 1, · · · , N} for some integer N and h := T/N. Let yh(tn) be approximation to y(tn). Assume that we have already calculated approximations yh(tj ), j = 1, 2, · · · , n and we want to obtain yh (tn+1) by means of the equation
yh (tn+1) =
where
aj,n+1
m−1 tnk+1 (k) |
hα |
|
P |
|
|
|
hα |
n |
|
|
|
|
|||||
X |
|
|
|
|
|
|
|
|
|
|
|
X |
|
|
|
|
|
=0 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
y0 + |
|
f tn+1, yh (tn+1) + |
|
|
aj,n+1f (tj , yn (tj )) |
|||||||||
k |
k! |
|
|
(α + 2) |
|
|
|
|
(α + 2) j=0 |
|
|
|
|
||||
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
(2.13) |
= |
nα+1 − (n − α)(n + 1)α, |
|
2(n |
|
|
if j = 0, |
|
n, |
|||||||||
(n |
− |
j + 2)α+1 + (n |
− |
j)α+1 |
− |
− |
j + 1)α+1, if 1 |
≤ |
j |
≤ |
|||||||
|
|
|
|
|
|
|
|
|
|
|
|||||||
|
1, |
|
|
|
|
|
|
|
|
|
|
if j = n + 1. |
The preliminary approximation yhP (tn+1) is called predictor and is given by
P |
m−1 tnk+1 (k) |
1 |
n |
||||
|
X |
|
|
|
|
X |
|
yh (tn+1) = |
|
k! |
y0 + |
(α) |
bj,n+1f (tj , yn (tj )), |
||
|
k=0 |
|
j=0 |
||||
where |
|
|
hα |
|
|
|
|
|
|
|
|
|
|
||
|
bj,n+1 = |
|
((n + 1 − j)α − (n − j)α) . |
||||
|
α |
Error in this method is
maxj=0,1,··· ,N |y(tj ) − yh(tj )| = O(hp),
where p = min (2, 1 + α).
(2.14)
(2.15)
(2.16)
3.The Model
A new 3D chaotic system proposed by Li et al. in [17] is
x˙ |
= |
a(y − x) + yz, |
|
y˙ |
= (c − a)x − xz + cy, |
|
|
z˙ |
= |
ey2 − bz, |
(3.17) |
where a, b, c and e are constants. We propose a modification in system (3.17) involving fractional order derivatives as below:
Dαx = a(y − x), |
|
||
Dαy |
= |
(c − a)x − xz + cy, |
|
Dαz |
= |
ey2 − bz, |
(3.18) |
where 0 < α ≤ 1. Now we discuss some dynamical properties of the system (3.18).
1.Symmetry. The system is symmetric about z-axis because the transformation (x, y, z) −→ (−x, −y, z) does not change the system.
Complimentary Contributor Copy
Dynamics of Fractional Order Chaotic Systems |
27 |
||||||
2. Dissipativity. Consider |
|
|
|
|
|
|
|
div(V ) = |
∂x˙ |
+ |
∂y˙ |
+ |
∂z˙ |
|
|
∂x |
|
∂z |
|
||||
|
∂y |
|
|||||
= c − (a + b). |
|
|
(3.19) |
The system is dissipative if c < a + b.
3.Non-generalized Lorenz system. The generalized Lorenz system (GLS) discussed in [20] is of the form
X˙ = |
A |
0 |
+ x |
0 |
0 |
|
1 |
X, |
(3.20) |
|
|
|
|
0 |
0 |
0 |
|
|
|
0 |
λ3 |
0 |
1 |
− |
|
|
|
||
|
|
|
|
0 |
|
where X = [x, y, z]T , λ3 and A is a 2 × 2 matrix with eigenvalues λ1, λ2 such that −λ2 > λ1 > −λ3 > 0. Since our system
X˙ = c−aa |
c |
0 |
|
+ x |
|
0 |
0 |
|
1 |
X + |
|
0 |
|
(3.21) |
|
|
− |
a |
0 |
|
|
|
0 |
0 |
0 |
|
|
yz |
|
|
|
|
|
−b |
|
0 |
|
− |
|
|
ey2 |
|
|
||||
0 |
0 |
|
1 |
0 |
|
is having different structure than the system (3), it is not a GLS.
4. Equilibrium points and their stability. The system (3.18) has three equilibrium
p p
points viz. E0 = (0, 0, 0), E1,2 = −b(a − 2c)/e, −b(a − 2c)/e, −a + 2c . The Jacobian matrix evaluated at an arbitrary point X = (x, y, z) is given by
J(X) = |
c |
|
a |
|
z |
c |
|
x . |
(3.22) |
|
|
|
−a |
|
a |
0 |
|
|
|||
|
|
− |
|
− |
|
|
− |
|
|
|
|
|
0 |
|
|
2ey |
−b |
|
The characteristic equation of J(E0) is therefore,
P = λ3 + (a + b − c)λ2 + (a2 + ab − 2ac − bc)λ + (a2b − 2abc) = 0. (3.23)
The following expression defines the descriminant of the equation (3.23).
D(P ) = −(3a2 − 6ac − c2)(a2 + b2 + bc − ab − 2ac)2. |
(3.24) |
The sign of D(P ) is same as the sign of −(3a2 −6ac−c2) =
√
(1 − 2/ 3)c)). The sufficient conditions for the stability of following theorem which are consequences of Theorem 2.3.
√
−(a−(1+2/ 3)c(a− E0 are discussed in the
p
Theorem 3.1. (a) If 2c < a < 13/3c then the equilibrium point E0 is asymptotically stable.
(b) If c = 0, b > 0 and a > 0 then E0 is asymptotically stable for α < 2/3.
Complimentary Contributor Copy