Chapter 6

Topological Methods Nonlinear Dynamics

Yalew Mekonnen Fenta

1
The nonlinear and linear system

 A linear system is a system that satisfies the superposition principle, which

states that the response of a linear system to any input is the sum of the
responses to each individual input. This means that if an input signal is
multiplied by a constant and then applied to the system, the output of the
system will also be multiplied by that same constant.
 A nonlinear system is a system that does not satisfy the superposition
principle, which states that the response of a linear system to any input is the
sum of the responses to each individual input. This means that if an input signal
is multiplied by a constant and then applied to the system, the output of the
system will not necessarily be multiplied by that same constant.
 Nonlinear systems can be represented using a mathematical model known as a
nonlinear differential equation, which describes how the system responds to an
input signal. Solving these equations can be challenging, as they may not have
closed-form solutions and may require the use of numerical methods.

2
 Nonlinear dynamics: concerns with dynamical systems or processes that are
inherently nonlinear. Nonlinear dynamical systems, describing changes in
variables over time, may appear chaotic, unpredictable, or counterintuitive,
contrasting with much simpler linear systems.
 The component form of n-th order system � = �(�) is:

(6.1)

 Linearity: The above system is linear if function � is a linear function. Functions

fi are linear combinations of the independent variables xi. Variables xi appear in
the first power only. No products, trigonometric, exponential, etc. functions of
xi are present.

3
 Nonlinearity: Any system that is not linear is nonlinear.
 Autonomous system: no explicit dependence on time t in �.
 Non-autonomous system: explicitly dependant on time t in �.
Example: Harmonic oscillator
Are the following systems linear or nonlinear? Autonomous or non-autonomous?
1. The equation of motion of the harmonic oscillator has the form:

(6.2)

where x is the displacement, m is the mass and k is the stiffness (of the spring).
 By introducing a variable exchange v = � (velocity) we rewrite the second-order
Eq. (6.2) as a system of first order ODEs

4
 (6.3)

 From here it is easy to see that this 2-D or second order system is linear and
autonomous.
2. The normalised dimensionless equation of motion for the mathematical pendulum
has the following form:
(6.4)
where θ is the angular displacement. By introducing a variable exchange ω = �
(angular velocity) we rewrite the second-order Eq. (6.4) as a system of first order
ODEs

(6.5)

 From here it is easy to see that this 2-D or second order system is nonlinear and
autonomous.
5
6
Linearization of system
 Figure below shows the Phase portrait of a 1-D problem or a first-order problem.
Fixed points are shown with the filled and unfilled bullets. Arrowheads show the
direction of flow.

 The problems of flow on a line given in the following form:

(6.6)

7
 where function f(x) can be linear or nonlinear.
 At times it is beneficial to simplify your problems in order to analyse them. This is
especially true in case of problems that are described by nonlinear differential
equations. Linearization is a tool for analysing the dynamics of nonlinear
systems.
Linearization of 1-D systems
 In mathematics, linearisation is finding the linear approximation to a function at a
given point. The linear approximation of a function is the first order Taylor
expansion around the point of interest. In the study of dynamical systems,
linearisation is a method for assessing the local stability and type of a fixed point
of a system of nonlinear differential equations or discrete dynamical systems.
 Let’s examine the dynamics of Eq. (6.6) close to its fixed point or points x∗. We
assume the solution is in the following form:
(6.7)
 where |η| << 1 is a small perturbation. The behaviour and change of solution x
over time thus is
(6.8)
8
 At the same time it holds that � = f(x). The combination of these results gives

(6.9)

 Neglecting O(η2) yields the linearisation of the system about fixed point x∗
(6.10)
 where s = f’ (x∗) is simply the slope of function f(x) evaluated at x∗

9
 Solution of Eq. (6.10) has the form

(6.11)
 where C is the constant of integration and η0 is a suitable initial amplitude. The
stability of this solution

 based on its behaviour is the following:

If s > 0, then the solution is exponentially growing (exploding). We say that
the solution dynamics and corresponding fixed point are unstable.

If s < 0, then the solution is exponentially decaying. Asymptotically

approaching a stable value (η(t) → 0 for t → ∞). We conclude that the
solution dynamics and corresponding fixed point are stable.

10
 Fig. (Top) Phase portraits of the original flow described by Eq. (6.6) and its linearised
counterparts given by eq. (6.10). (Bottom) Aforementioned linearised solutions shown on
their respective phase portraits.
11
Phase Portrait

 Phase Portrait is a useful graphic tool to understand the stable and unstable
behaviour of equilibrium points of a nonlinear systems.

 It is a graphical method for studying second-order systems by

providing motion trajectories corresponding to various initial conditions.

then examine the qualitative features of the trajectories.

finally obtaining information regarding the stability and other motion patterns of
the system.

12
 Bifurication in 2D
 Here we present a selection bifurcations that most commonly occur in practical
applications. Bifurcations in two-dimensional systems are classified as follows:
CASE I Bifurcations of fixed points
A) Bifurcations at λ1 = 0 or λ2 = 0 where λ1,2 are the system eigenvalues.
1) Saddle-node bifurcation
2) Transcritical bifurcation
3) Pitchfork bifurcation
∗ Supercritical pitchfork bifurcation
∗ Subcritical pitchfork bifurcation
B) The Hopf bifurcations, bifurcations at λ1,2 = ±iω
1) The supercritical Hopf bifurcation
2) The subcritical Hopf bifurcation

13
CASE II Global bifurcations of closed orbits
A) Saddle-node coalescence of cycles (accompanied by the subcritical Hopf)
B) SNIPER (saddle-node infinite period bifurcation) also called SNIC (saddle-
node in invariant cycle bifurcation)
C) Homoclinic bifurcation or saddle-loop bifurcation

Saddle-node Bifurication
 The saddle-node bifurcation is the basic mechanism for creation and destruction of
fixed points.

We have seen the cases

for the equation:

14
 A similar general behaviour, i.e. bifurcation dynamics, happens for the system
with inverted parabola

 The bifurcations of fixed points have analogs in two dimensions,

15
Phase portrait of saddle-node bifurication

16
Transcritical bifurcation
 The normal form of the transcritical bifurcation is given by

 where r is the control parameter.

17
 In a transcritical bifurcation the fixed points are not destroyed whereas their
stability is changed.

18
Pitchfork bifurcation
 The pitchfork bifurcation occurs in systems with symmetry. It involves mergers
or splitting apart of fixed points and changes in their stabilities. Normal form is
given by

Supercritical Pitchfork bifuricationphase portrait

19
 Supercritical pitchfork bifurcation
 In the case of supercritical pitchfork bifur_x0002_cation the bifurcating pitchfork
branches are stable fixed points.

 Case a = 0 features algebraic decay or critical slowing of the flow as it approaches

the fixed point.
20
Subcritical pitchfork bifurcation

 In the case of subcritical pitchfork bifurcation the bifurcating branches are

unstable fixed points.

 Case a = 0 features algebraic decay or critical slowing of the flow near the fixed
point.
21
The supercritical Hopf bifurcation
 The normal form of the supercritical Hopf bifurcation, given in polar coordinates,
is the following

22
The subcritical Hopf bifurcation
 The normal form of the subcritical Hopf bifurcation given in polar coordinates is
the following

23
Bifurcations of closed orbits

24
The End

25