Construction of Phase Portraits

I. Analytical method
II. The method of isoclines
III. Delta Method
 (1) Analytical Method
• Two Ways:
(a) Integrate (if possible) the equation of slope of the
trajectories given by: dx 2 f ( x1 , x 2 )

dx 1 x2
• such that x 2   ( x 1 )
(b) Solve for x 1 and x 2 from state equations as
function of time, t.
• Then eliminate time variable t from the expression
for x 1 and x 2 .
• Not always possible.
• Example: If differential equation of a linear system
is given as: d 2 y
2
  y  0
2

dt
• Then state equation are:
dx 1
 x2
dt
dx 2
   2 x1
dt
• Equation of slope of state trajectory:
dx 2  2 x1
 
dx 1 x2
• Cross-multiplying: x dx   2 x dx  0
2 2 1 1
• Integrating, one gets: x 22 x 12
 2
 c
2 2
• On simplification:
2
 x2 
   x 2
 c1
 
1

• Example 2: Consider a non-linear closed-loop

system as shown in following figure. Obtain the
State equations and equation(s) for state-
trajectories in phase plane.
r e 1 u 1 y
1
 1
s ( s  1)
 1

r  constant
• Solution: From plant dynamics:
.
.. .
y y  u (1 )
• For non-linear element , input-output relation is
given by: u  f ( e ) (2)
where f ( e )   1 for e  1
 0 for  1  e  1
  1 for e   1
• From comparator dynamics: e  r  y
• Differentiating above equation:
.
y  e
.

.. ..  (3)
y  e 
 .. .

• Putting (3) and (2) in (1), one gets: e  e   f ( e )

.
• Taking state variables as: x1  e and x2  e
.
• State equation can formed as: x1  x2
.
x2  x2  f (x1)
where f ( x 1 )   1 for x 1  1
 0 for  1  x 1  1
  1 for x 1   1
• Slope equation can be given as:
dx2 x2  f (x1)

dx1 x2
• Cross-multiplying and rearranging:
 x2
dx2  dx1 (4)
x2  f (x1 )
• Using substitutions method, Let x2  f ( x1)  z
then dx2  dz and x2  z  f ( x1 )
• Equation (4) become:
z  f ( x1 )
dz   dx1
z
• Integrating both sides:
z  f ( x1 ) ln z   x1  c
• In terms of x2, we get equation of trajectories as:
x2  f ( x1 )  f ( x1 ) ln  x2  f ( x1 )    x1  c

• Find the value of c using the initial value of x1 and x2

• Equations of trajectories are given as:
Region I x 1  1
 x2  1 
 x1  x1 (0)     x2  x2 (0)   ln  
 x2 ( 0)  1 
Region II   1  x1  1
x1
 x1 (0)     x2  x2 ( 0) 
Region III x 1
 1
 x2  1 
 x1  x1 (0)     x2  x2 (0)   ln  
 x2 ( 0 )  1 
 (2) Isoclines Method
• If state equation is given as:
dx 1
 f 1 ( x1 , x 2 ) (1)
dt
dx 2
 f 2 ( x1 , x 2 ) (2)
dt
• Slope of the state trajectories is given by:
dx 2 f 2 ( x1 , x 2 )

dx 1 f 1 ( x1 , x 2 )
• For a constant slope, say N, then:
dx 2 f 2 ( x1 , x 2 )
  N
dx 1 f 1 ( x1 , x 2 )
S ( x1 , x 2 )  N (3)
• Equation (3) define as a curve in x1-x2 plane along
which state trajectories of (1) and (2) have slope
N.
• Whenever a trajectory of (1) and (2) cross the
curve defined by (3) , slope of trajectory at the
point of intersection point will be N.
• Equation (3) gives the locus of constant slope of
the trajectory .
• Locus of point where the phase trajectories have a
constant slope.
• This equation (3) gives the equation to the family
of isoclines.
• Procedure:
(a) For different value of N , draw the isoclines curves
sufficiently so to fill the x1-x2 plane with isoclines.

(b) along each isoclines, draw small line segment

having slope , N . Line segments are parallel. Their
direction are determined by the sign of N.

(c) Starting from a given initial point (x1(0), x2(0)), one

can construct state trajectory by moving in
direction of short line segments from one isocline to
next.
• To understand the concept of isoclines method:
• Let us take a simple linear system given by:
d2y dy
2
 2  n   n y  0
2

dt dt
dy
• Taking the state variables as: x1  y and x 2 
dt
• Taking   0 .5;  n  1 , we get state equations as:
dx 1
 x2
dt
dx 2
  x 2  x1
dt
• Slope equation becomes: dx ( x 2  x1 )
2
 N
dx 1 x2
• Equation of locus of points having same “N”
(isoclines) is given by: 1
x  x
(1  N )
2 1

x2 N  1
N 0
N  2
N  3

N  
x1
D) Simple steps to find approximately the
next point in the state trajectories:
L1
• If P1 is a point in state trajectory and L1
and L2 be two isoclines for N = N1 and P1 L2
N=N2 respectively.

• Then From P1 , draw a line parallel to line A

segments in L1 , meeting L2 at A.
B
• Then From P1 , draw a line parallel to line
segments in L2 , meeting L2 at B.

• Locate the midpoint of AB as the next

point P2 of state trajectory.
• Repeat the above steps from P2 to find P3.
 (3) Delta Method
• Convenient when the second order system is likely
to be oscillatory with low damping.
• Method consists in approximating the trajectory
with short arcs of circles. .. .
• The differential equation: y   ( y , y )  y  0
is arranged in form the state equations:
.
x1  x 2
.
x 2    x1   ( x1 , x 2 ) 
• Slope equation:
dx 2   x1   ( x1 , x2 ) 

dx1 x2

• Cross multiplying: x2 dx2   x1   ( x1 , x2 )  dx1  0

• If the function  ( x1 , xis2 sufficiently

) small ,it can be taken as
constant for small changes in operating points
• Then integrate the equation:
x2 dx2   x1   (.)  dx1  0
• We get:
x   x1   (.)   R
2 2 2
2

• This is equation to a circle with center at   and 0

(.),radius
=R
• Part of the trajectory can be drawn as circle with the
centre at    and
, 0  radius R.
• Construction requires evaluation of  at different
points in state plane.
• Drawing arcs of circles from the centers at
   , 0 
and
appropriate radius.
• Following procedure can be used for construction of
the trajectory:  ( x1 , x2 )
 ( xthe
1) Identify
1 , x 2
term
) tox be
1 or include
x 2
in function
2) Plot vs. P1 , ( x1 , x2 ) be.
as case
0 may
0

3) From a given initial1 state project lines

to determine and set this   length
1 , 0  against the
negative real axis locating .
4) With centre at    1 , 0  and radius determined by the
position of the initial state, drew an arc of a circle
forming a part of trajectory.
5) Repeat Steps 3 and 4 at the new point on the trajectory
and complete the trajectory repeating these steps.

Figure for Delta method of Phase plane trajectory

construction
• Example of a hydraulic speed control system :
.. .
e  0 . 33 e  (1  0 . 58 e ) e  0
2

• Defining the error, e(t) and its derivative as the state

variables:
.
x1  x 2

 
.
x 2   x 1  0 . 33 x 2  0 . 58 x 3
1
•  can be identified as 0 .58 x 3  0 .33 x which
1 2
can be split into two terms:
 1 ( x1 )  0 .58 x 3
1 and  2 ( x 2 )  0 .33 x 2
•  1 ( x1 ) is plotted against x1 .
•  2 ( x 2 ) is plotted against x 2 .
• From an initial condition point, values of  2(1) ( x 2 )
and  1(1) ( x1 ) are measured.
• These are laid out towards the negative x1 axis as
 (1 )    (1) (1 )

• With point    (1) , 0  as centre and radius equal to

1 2

distance between the centre and initial condition

point, an arc of a circle is drawn .
• The procedure is repeated at other points as well to
completer the trajectory. .
 Symmetry in Phase Plane Portraits
.. .
y  f ( y, y)  0
•Consider the differential equation :

•In phase variable form, it become:

.
x1  x 2
.
x 2   f ( x1 , x 2 )
• Slope of phase trajectories in Phase Portraits:

dx 2  f ( x1 , x 2 )

dx1 x2
(a) Symmetry about the x1 axis:
• Slope must be equal but opposite in sign for x2 > 0 and
x2 < 0 for all x1 .
• Possible if f (x1 ,,x2 ) = f (x1 ,, － x2 ) .
• i.e. f (x1 ,,x2 ) must be an even function of x2
• If phase portrait is symmetrical about x1 axis, only
difference between the phase plots in the upper and
lower half is in direction of motion.
• Motion of the representative point on the upper half will
be the right and in lower half, it will be to the left with
respect to x1 axis
(b) Symmetry about the x2 axis:
• Slope must be equal but opposite in sign for x1 > 0 and x1 < 0 for all x2 .
• Possible if f (x1 , x2 ) = － f ( － x1 , x2 )
• i.e. f (x1 , x2 ) must be an odd function of x1

(c) Symmetry about both the axes:

• Both the above conditions of (a) and (b) must be true, simultaneously.
• Possible if f ( － x1 , x2 ) = － f ( x1 , － x2 ) .
 Limit Cycles
• A limit cycle is an isolated closed trajectory.
• Discovery by H. Poincare

• Examples:
● Heart beat, pacemaker, neurons, daily rhythms in
body temperature, chemical reactions, ...
• A limit cycle is an isolated closed trajectory;
• this means that its neighboring trajectories are not
closed.
• they spiral either towards or away from the limit cycle.
• Thus, limit cycles can only occur in nonlinear systems.
• In a linear system exhibiting oscillations closed
trajectories are neighbored by other closed trajectories

• A stable limit cycle is one which attracts all neighboring

trajectories.
• A system with a stable limit cycle can exhibit self-
sustained oscillations – most of the biological processes
of interest are of this kind.
• Neighbouring trajectories are repelled from
unstable limit cycles.
• half-stable limit cycles are ones which attract
trajectories from one side and repel those on the
other.
• Trajectory of Limit Cycle has to be both closed,
indicating the periodic nature of motion, and
• Isolated , indicating the limiting nature of the cycle
with neighboring trajectories converging to or
diverging from it.
• We have seen examples of linear and non-linear
systems which have a family of periodic solutions.
• These occur near any centre of the system.
• Any such periodic trajectory is called a cycle. There
is a type of cycle which is isolated i.e. it is not part
of a family of cycles.
• We are interested in the question of stability of such
cycles.
• That is, do trajectories which are near to a cycle,
approach it or leave it as t→∞?
• If all trajectories which are sufficiently close to a
cycle, converge to that cycle as t→∞,
• we call the cycle an asymptotically stable limit cycle
of the system.
• If no trajectories which are sufficiently close to a
cycle, diverge from that cycle as t→∞,
• we call the cycle an unstable limit cycle of the
system.
• If some trajectories which are sufficiently close to a
cycle, converge to that cycle as t→∞ and others
diverge from that cycle,
• we call the cycle a semi-stable limit cycle of the
system.
• Are inherently a non-linear phenomenon
● What about .closed orbits in linear systems?
X  AX
They can have solutions that
are closed orbits!

● Linearity implies: if x(t) is a solution so is c x(t)

● Amplitude is determined by Initial conditions

● In contrast: in limit cycles in nonlinear systems, the

structure of the system determines the amplitude
and shape of the limit cycle.
• Oscillation: A system oscillates when it has a
nontrivial periodic solution.
x ( t )  x ( t  T ); for all t  0

• The linear oscillation is not practical because :

• It is not structurally stable.
• Infinitesimally small perturbations may change the
type of the equilibrium point to a stable focus
(decaying oscillation) or unstable focus (growing
oscillation)
• The same problems exist with oscillation of
nonlinear systems due to a center equilibrium point
(e.g., pendulum without friction)
Van der Pol Oscillator

  y
.. .
y  1 y 2
y  0
 Equation arose in connection with
nonlinear electrical circuits used.
• Damping term
Causes large amplitude oscillates
to decay, but pumps them back if
they become too small
● Van der Pol Oscillator has a unique
stable Limit cycle for each  > 0
 Poincaré-Bendixson Theorem
• As closed orbits are of much interest there
should be a method to establish that there
exists such an orbit, and there is: Poincaré-
Bendixson Theorem:
• Suppose that:
1. R is a closed bounded subset of the plane;
2. dx/dt = f(x) (state equations) is a continuously
differentiable vector field on an open set
containing R;
3. R does not contain any Fixed Points (Singular
points);
4. There exists a trajectory C that is confined in
R, in the sense that it starts in R and stays in R
for all future time.
• Then either C is a closed orbit, or it spirals
towards closed orbit as time tends to infinity .
• So, R contains a closed orbit.
• Bendixson‘s Criteria for the existence of the limit
cycle:
• Bendixson has shown that a sufficient condition for
the absence of the limit cycle or any sort of closed
trajectory within a region is that ,
• Given the state equation of a non-linear system as:
.
x1  f1 ( x1 , x 2 )
.
x 2  f 2 ( x1 , x 2 )
• Then on the boundary of the region:
 
f1 ( x1 , x 2 )  f 2 ( x1 , x 2 )  fixed sign
 x1 x2
Proof: Slope equation of trajectory is given as:

dx 2 f 2 ( x1 , x 2 )

dx 1 f1 ( x1 , x 2 )
on cross multplying :
f 1 dx 2  f 2 dx 1  0

• This equation gives relation that must exit along any

solution trajectory in the state plane.
• Gauss Theorem states that a surface integrals over a a
complete domain R is related to a line integral around
the closed path C which is boundary of region:
    
R   x f1   x f 2  dx 1dx 2    f dx
1 2  f 2 dx 1 
 1 2  c

• If a limit cycle C is assumed to be bounding the

region R, Then,
• The RHS of the Gauss Theorem is zero since the
integral is zeros everywhere on C.
• Thus , the LHS must be zero
• Consequently,
 
f1  f2
 x1 x2
must change sign with R.
 Example:
• Consider system represented by the following
differential equation:
d y 2 10  dy   dy
2

 0 .1     y y 0
2

dt 2  3  dt   dt
 
• State space representation cab be given as:
dx 1
 x2
dt
dx 2  10 2 
  0 .1   x 2   x 2  x1  x 12

dt  3 
so , f 1 ( x1 , x 2 )  x 2 ;
  10 2 
and f 2 ( x1 , x 2 )   0 . 1   x 2   x 2  x1  x 1
2

 3 

Thus , f 1 ( x1 , x 2 )  0
 x1

and f 2 ( x1 , x 2 )  0 . 1  10 x 2 2

x2
 
 It is a apparent that f1 ( x1 , x 2 )  f 2 ( x1 , x 2 )
 x1 x 2
does not have fixed sign.

 By Bendixson’s criteria, limit cycle may exit.

 Concept of Feedback Linearization
• Consider a class of non-linear system in the form of:
d
x  f ( x )  G ( x )u
dt
y  h( x)
• If there exit a state feedback control :
u   ( x )   ( x )v
z  T ( x)
• and a change of state variables, ,
• That transform the non-linear system into an
equivalent linear system.
• Then the classical linear control techniques can be
applied to resulting closed loop system.
• Feedback Linearization can possible in two forms:
(1) Input-state linearization
• In this the full state equation is linearized
(2) Input-Output Linearization
• In this case emphasis is given on linearizing the
input-output map from u to y, even if the state
equation is only partially linearized.
Input-state linearization
• Consider example of Pendulum system, having
following differential equation:
d 
2
d
2
  a sin   b  cT
dt dt
• Where T is applied torque and  is angle subtended by
rod and the vertical axis.
• Objective is to stabilize the pendulum at an angle
  .

• Choosing the state variables as x1     ; x 2  

• So that desired equilibrium point is located at the origin
of the state space
• dx
The state equation
dx is obtain as:
1
 x2 ; 2
  a  sin( x1   )   bx 2  cT
dt dt
• For system to maintain equilibrium at the origin , torque
must have steady state component Tss, that satisfies
0   a  sin    cT ss
• Define a control variable u as: u  T  T ss
• Rewrite state equation:
dx 1
 x2 ;
dt
dx 2
  a  sin( x 1   )  sin    bx 2  cu
dt
• From inspection of the state equation shows that if
we choose u as below to cancel nonlinear term:
a v
u   sin( x1   )  sin   
c c
• results in:
dx 1 dx 2
 x2 ;   bx 2  v
dt dt
• Thus stabilization problem for non linear system has
been converted to stabilization problem for a
controllable linear system
• We can now proceed to design a stabilizing state
feedback control law as:
v  k 1 x1  k 2 x 2
• To locate the eigen values of the closed loop system
in the left half of s-plane.

dx 1 dx 2
 x2 ;  k 1 x1  ( k 2  b ) x 2
dt dt
• The overall state feedback control law is given by:
 a 1
u   sin( x1   )  sin     k 1 x1  k 2 x 2 
c c
Generalization of non-linearity cancellation
 Not to except that we can cancel nonlinearities in
every nonlinear system.
 Certain structural property of the system that allow
to perform such cancellation.
 ( x)
 To cancel a nonlinear term by( xsubtraction,
)
u   ( x)
the control u and the nonlinearity must
appear together as a sum  ( x )
 To cancel a nonlinear term  ( xby ) division,
 ( x )u
control u and the nonlinearity must appear
together as a product .
• If the matrix  ( x ) is non singular in the domain of
interest , then it can be cancelled by u   ( x ) v
where  ( x )   ( x ) .
1

• Ability to use feedback to convert a nonlinear state

equation into a controllable linear state equation by
cancelling nonlinearities requires the nonlinear
state equation to have the structure:
d
x  Ax  B  1
( x ) u   ( x ) 
dt
• Where A = n X n , B = n X p , pair (A,B) is controllable
• Functions
 ( x) : R 
 R ;  ( x ) : R 
n p
 R n pXp
 are defined in the domain D x  R n that contains
the origin.
• If the state equation takes the form as above, then
one can linearize it via the state feedback law:
u   ( x )   ( x )v
• To obtain the linear state equation:
d
x  Ax  Bv
dt
• For stabilization, one design v  Kx such that A BK
is Hurwitz . The overall nonlinear stabilizing state
feed back control law is:
u   ( x )   ( x ) Kx
• For Pendulum stabilization problem having the state
equation as:
dx 1
 x2 ;
dt
dx 2
  a  sin( x1   )  sin    bx 2  cu .
dt
• System state equation can be rearranged in matrix
form as:
 x  0 1   x  0  a
.

 . 
1
 
1
   c u   sin(x1   )  sin  
 x2  0  b  x2  1  c 
• Comparing with structure,
d
x  Ax  B  1
( x ) u   ( x ) 
dt
• we get
0 1  0  1
A    ; B   ;  (x)  ;
0  b 1  c
a
 (x)   sin( x 1   )  sin  
c
• For Linearization, control law used :
u   ( x )   ( x )v
• If a non-linear state equation does not have the
structure as above for one choice of state variables,
it might do so for
.
another choice. .
• For example: x1  a sin x 2 ; x 2   x12  u
• One can cannot cancel nonlinear term a sin x2
• But if change variable by the. transformation:
z1  x1 and z 2  x1  a sin x 2
• So new state equation:
. .
z1  z 2 ; z 2   ax cos x 2  au cos x 2
2
1
• Nonlinearities can be cancelled by the control:
1
u x 
1
2
v
a cos x 2
• State equation in new coordinates (z1, z2) can be found by
transforming to express (x1,x2) in terms of (z1,z2).
• Transformed state equation:
  z2  
 
. .
1
z1  z 2 ; z 2   a cos  sin    z 1  u
2

  a 
z  T ( xto) z-coordinates is
• Transformation from one x-coordinates
1
done through
1
T (.)
x  T An
• The map T must be invertible. ( zinverse
) z  T (Dx )
map
must exit such that for all
where Dx is the domain of T.
 Since the derivatives of z and x should be continuously ,
both T(.) and inverse of T be continuously differentiable.
 A continuously differentiable map with a continuously
differentiable inverse is known diffeomorphism.
 Portion left in First Module
Computer simulation:
• to find the response of non-linear system taking
different initial conditions and inputs.
• Draw the phase plane portrait of a non-linear
system using computer and a linear system and
comparing with that obtained from delta and
isoclines methods.
• Concept of inverse control and model predictive
control