INDIAN INSTITUTE OF TECHNOLOGY ROORKEE
PHASE POTRAITS
PRESENTATION III
PRESENTED BY:-
MAJ MIHIR GHOOI
DEEPAK DUBEY
MOREIRA FERNANDES
INTRODUCTION
Phase portraits
Geometric representation of trajectories of a dynamical sys in ph
plane.
Depicts system trajectories with arrows and stable states with dots.
Tool in studying dynamical systems.
Gives info regarding chosen parameter value.
Graphical tool, visualizes how solutions of a given system of
differential equations behave in the long run.
2
DEFINITIONS
Dynamical system is a system in which function describes time
dependence of a point in a geometrical space.
Autonomous system is a system of ordinary differential equations that
does not explicitly depend on time.
Critical points on the graph of a function are points where the derivative
is zero.
Locus in x1-x2 plane of solution x(t) is called Trajectory.
Family of phase plane trajectories for different initial conditions is called
Phase Portrait of the system.
3
METHODS TO FIND PHASE POTRAITS
VECTOR FIELD METHOD
ISOCLINE METHOD
DELTA METHOD
PELLS METHOD, etc
4
VECTOR FIELD METHOD
= f(x) = [f1(x); f2(x)] , = (1, 2)
To each vector (x1, x2) a vector
known as vector field (f1, f2) is
associated.
Example
1= 2x12
2= x2
1= x2
2= -10sinx1
5
ISOCLINE METHOD
Means same slope.
Consider
S(x) = dx2/dx1= f2(x1; x2)/f1(x1; x2)
An isocline is defined as s(x) = c
That is S(x) = dx2/dx1= f2(x1; x2)/f1(x1; x2) = c
All points on the curve f2 =c f1 have the same tangent slope
c.
6
ISOCLINE METHOD ALGORITHM
Draw curve f2(x) = cf1(x) with
different values of c.
Draw small lines with slope c on
each curve.
Direction of the line at a point will be
given by the sign of f1 and f2 at the
point.
These small lines of all isoclines
constitute the phase portrait.
7
ISOCLINE METHOD ALGORITHM
Pendulum without Friction: 1= x2 2= -sin(x1)
Pendulum with Friction: 1= x2 2= -0.5x2-sin(x1)
8
USING PHASE POTRAITS
Consider a Linear system = Ax, where matrix A is a 2X2
diagonal matrix.
1 = -6x1
2 = -3x2
1 = 6x1
2 = 3x2
9
Eigen values
Gives us info about how fast the rate will be
Gives us idea about the nature of the convergence
I-A=0 gives the different Eigen values
Eigen vector
Gives us idea about how the data will be spread in any system
Gives us information about axes along which the system tends to
stretch /compress / flip
For a 22 matrix AX=x will give us Eigen vector
Jordan form
; ;
10
Modal coordinate vs XY plane
Modal coordinate
Datas are transformed into a coordinate system where their eigen
vectors represents two orthogonal axes
Matrix A is transformed into Jordan form (replacing x=Mz, M=[v1 v2])
Gives clear idea about the comparative magnitude of the eigen
values
Mathematically we can see its movement
XY plane-
More realistic view of how our datas converge/diverge
Can be directly drawn, no need of modal coordinate
Both of them gives idea about stability of the system
11
Phase portrait in Modal coordinate
Transformed system results in following equation
Solution for initial condition (z10,z20)
Replacing time of first equation into second results
Vary initial condition to vary C thus we get phase portrait.
Phase portrait depends on sign of .
12
Procedure to draw phase portrait in XY
plane(2nd order)
Find critical points:
Eg: x=4x-3y ,y=6x-7y have critical point at (0,0)
construct a phase plot (y vs x)
find eigen values and eigen vector of the system equation
eigen values are (-5 and 2) and corresponding eigen vectors are [1;3] and
[3;2] and draw corresponding vector axes
if eigen value is ve then solution
will grow towards critical point and
if it is positive then soln will flow
away from the critical point
(ie diverging)
decide the direction of flow of the flow
on the axes corresponding to eigen
vectors. And all other soln follow the
direction of the flow of the vector axes.
13
Different possible cases
real and different eigen values
both positive
both negative
one positive and one negative
real and same eigen values
both positive
both negative
both zero (doubly degenerate form)
complex eigen values where real part is
positive
negative
zero
14
Real and different eigen values
both positive
A= = v1= v2=
15
Real and different eigen values
both negative
A= = v1= v2=
16
Real and different eigen values
One negative and one positive
4 0 4 5 0
A= = v1= v2=
2 -1 -1 2 1
17
Real and different eigen values
One negative and one positive
1 .5 0 1 1
A= = v1= v2=
2 1 2 -2 2
18
Real and same eigen values
Both are positive
2 -3 3 3
A= = v1= v2=
1/3 4 3 -1
19
Real and same eigen values
Both are negative
-7 1 -5 1
A= = v1= v2=
-4 -3 -5 2
20
Real and same eigen values
Both are zero
1 -1 0 1
A= = v1= v2=
1 -1 0 1
21
Real and same eigen values
Both are same with k=0
-3 0 3 1 0
A= = v1= v2=
0 -3 3 0 1
22
Complex eigen values
Real part -ve
-1 3 -2+3i
A= =
-3 -4 -2-3i
23
Complex eigen values
real part is +ve
A= =
1 3 2+3i
-3 4 2-3i
24
Complex eigen values
Real part is zero
0 1 2.24i
A= =
-5 0 -2.24i
25
26
Local Behaviour of Nonlinear Systems
Perturbations and reaction of the system around equilibrium points determines
the degree of validity of the behaviour analysis through linearization
Example:
Consider the linear perturbation case:
A A + A, where A is the small perturbation
Positive (Negative) eigenvalues of A remain positive (negative) under small
perturbations.
Therefore, node or saddle point or focus equilibrium point remains the same under
small perturbations. There are called structural stable
27
Local Behaviour of Nonlinear Systems
For eigenvalues on the j axis no matter how small perturbation is,
it changes the sign of eigenvalue.
Different Perturbations:
1 1 1
1 0 0
28
Qualitative Behaviour Near Equilibrium
Points
Qualitative behaviour of second-order nonlinear system can be
investigated by
1. Generating phase portrait of system globally by computer
programs
2. Linearization of the system around equilibrium point
29
Linearization of the system around equilibrium
point
Let (0) = (10 , 20 ) are equilibrium points of
. 1 = 1 (1 , 2 )
(1)
. 2 = 2 (1 , 2 )
1 , 2 are continuously differentiable about (10 , 20 );
Since we are interested in trajectories near (10 , 20 ) , define
1 = 10 + 1 Where 1 , 2 are small perturbations from
2 = 20 + 2 equilibrium point
Expanding (1) into Taylor series and neglecting the higher order terms we get:
. 1 = 11 1 + 12 2
.. 2 = 21 1 + 12 2
30
Linearization of the system around equilibrium
point
In vector form:
1 1
11 12 1 2 Jacobian Matrix
= 21 = 2 2
21
1 2
The trajectories of the nonlinear system in a small neighbourhood of an equilibrium
point are close to the trajectories of its linearization about that point:
If the origin of the linearized state equation is a stable (unstable) node, or a
stable (unstable) focus or a saddle point
Then in a small neighbourhood of the equilibrium point, the trajectory of the
nonlinear system will behave like a stable (unstable) node, or a stable (unstable)
focus or a saddle point.
31
Example 1.: Tunnel Diode Circuit
Global Phase Portrait generated by computer Program
32
Example 1.: Tunnel Diode Circuit
Linearization around the equilibrium points
1 = 0.5( 1 + 2 )
. 2 = 0.2(1 1.52 + 1.2)
1 = 17.761 103.7912 + 229.6213 226.33114 + 83.7215
.5(1 ) .5
=
.2 .3
1 = 17.76 207.581 + 688.8612 905.2413 + 418.614
Eigen values of the matrix at equilibrium point:
For Q1=(0.0626, 0.7582) 1 = -3.57, -0.33
For Q2=(0.2854, 0.6098) 2 = 1.77, -0.25
For Q3=(0.8844, 0.2104) 3 = -1.33, -0.4
33
Example 1.: Tunnel Diode Circuit
Linearization around the equilibrium points
3.598 0.5 1.82 0.5 1.427 0.5
1 = 2 = 3 =
0.2 0.3 0.2 0.3 0.2 0.3
Eigen values of the matrix at equilibrium point:
1 = -3.57, -0.33
2 = 1.77, -0.25
3 = -1.33, -0.4
34
Example 1.: Tunnel Diode Circuit
Global Phase Portrait generated by Linear System
35
Example 1.: Pendulum
Global Phase Portrait generated by computer Program
1 = -0.25 j0.97
2 = -1.28, 0.78
36
Linearization of the system around equilibrium
point
Exception: If the Jacobian matrix has eigenvalues on j, then the qualitative
behaviour of nonlinear system near the equilibrium point could be quite distinct
from the linearized one.
37
Numerical Construction of Phase Portrait
Step 1. Find the Equilibrium points
Step 2. Determine the type of isolated equilibrium point via
linearization
Step 3. Draw the trajectories:
3.1: Select bounding box
3.2: Select initial conditions
3.3: Calculate the trajectories
38
39
