Local stability, Limit

Cycles
& Poincare Bendixson
Theorem
By Priyanshi (14363) and Lavisha (14369)
(BSc. Physical Science with Chemistry)
Local Stability (2- D)
In systems like population models, epidemics…etc , we often deal with differential
equations that describe how a system changes with time.
A fixed point (or equilibrium point) is a point where the system does not change .

Given system: x• = ax + by
y• = cx + dy

( a b)
J= c d

We have to determine nature and stability of fixed point at (x,y) = (0,0)

Trial Solution
X = αeλt
y = βeλt
On putting these values in system equations
(a – λ)α + bβ = 0
(d – λ)β + cα = 0
For non-zero α,β :
( a – λ) ( d – λ) – bc = 0
Characterstic Equation :
Λ2 – (a+d) + ( ad – bc ) = 0

Roots are λ1, λ2 where

2 λ1 = a+d + {(a-d)2 + 4bc}1/2
2 λ2 = a+d - {(a-d)2 + 4bc}1/2
Here λ1 and λ2 are eigenvalues of the matrix:

( a b)
J= c d

Case 1: λ1 and λ2 are real and distinct

Assume b is not equal to 0 :
x = b (C1e* λ1t + C2e* λ2t)
Y = ( λ1 – a)C1e* λ1t + ( λ2 - a) (C2e* λ2t)
C1 and C2 are constants
So, equations : (λ2 – a)x - by = b (λ2 - λ1) C1e* λ1t
(λ1 – a)x - by = b (λ1 – λ2) C2e* λ1t

(λ2 – a)x – by , (λ1 – a)x – by

These are the boundary lines in phase plane.

lines in phase plane

Case 3: λ1 and λ2 < 0
t ∞
x 0
y 0

All trajectories move toward the origin.

Hence, a fixed point of this type is known as a Stable Node .
All arrows in the phase plane point inward.
Example - Rolling of a ball in the bowl, it doesn't matter from where
it was started. It will stop at the centre.

Stable Node

Case 4: λ1 and λ2 > 0

 t ∞
x ∞
y ∞
All trajectories move outward the origin.
Hence, a fixed point of this type is known as Unstable Node.
Example : On pushing a ball on the top of the hill it will roll away.

Case 5: λ₁, λ₂ with λ₁ > 0 and λ₂ < 0

Trajectory start from the origin and moves away from the origin.
 the fixed point will be known as Saddle.

Saddle point
Case 6: λ₁, λ₂ = α+- iω ( complex conjugate form )
(a-d)2 + 4bc < 0
Eigen values will complex.
Is λ1 and λ2 are complex conjugate but not imaginary.
Then the fixed point will be known as spiral.

Pure Imaginary Values : (a+ d) = 0

Suppose a+d = 0, λ1 and λ2 will be purely imaginary.

−ω∗2 A∗2
Equation of trajectory : cx2 + ( d-a)xy – by2 = b

 Then the fixed point will be known as Centre.

 And the motion of trajectories will be Closed.

Centre
Exponential Spiral Trajectories ( a+ d) ≠ 0
 If a + d ≠ 0 , trajectories will be called Spirals.
 If a + d < 0 , Stable Focus ( Spirals towards origin)
 If a + d > 0 , Unstable Focus ( Spirals outwards origin)

Stable and unstable focus

Case 7: λ1 = λ2 Repeated Eigen Values

(a - d)² + 4bc = 0
Fixed point will be known as Node.

 a < 0 , Stable node

 a > 0, Unstable node
Real life Significance local Stability-
1. Population Control
 Helps check if animal or human populations stay steady or change too much.
 Example: Rabbit population in a forest stays around the same number if the system is
stable.
2. Disease and Epidemics
 Shows whether a disease will disappear or keep spreading in a community.
 Example: Vaccination helps keep the number of infected people low and stable.

Example 1).
X• = 6x + 12y
y• = 3x + y
For fixed point/equilibrium point / critical point :
X• = 0
y• = 0

So, 6x + 12y = 0
3x + y = 0

Now, (6 121 )
J= 3

( 6-λ) (1- λ) – 12.3 = 0

6 - 6 λ – λ + λ2 – 36 = 0
Λ2- 7 λ – 30 = 0
 7 ± √ 49−120
λ=
2

λ1, λ2 = 10 , -3 ( which are real and opposite sign )

saddle point

Example 2).
X• = 5x + 2y
y• = 2x +2y
For fixed point / equilibrium point / critical point :
X• = 0
y• = 0

So, 5x + 2y = 0
2x +2y = 0

Now, ( 5 2)
J= 2 2

(5-λ) (2- λ) – 2.2 = 0

10 - 7 λ – λ + λ2 – 4 = 0
Λ 2- 7 λ – 6 = 0
7 ± √ 49−24
λ=
2

λ1, λ2 = 1 , (which are real and positive)

unstable node

Example 3).
X• = 5x + 4y
y• = 9x +0y
 (5 4 )
J= 9 0

λ1, λ2 = 9, -4 (one negative and another one is positive )

saddle point

Example 4).X• = -3x - y

y• = -2x -4y

(−3
J = −2 −4
−1
)
λ1, λ2 = -5, -2 ( which are negative )

stable node

Example 5). X• = -1x - 5y

y• = 5x -1y

J= 5 (−1 −5
−1 )
λ1, λ2 = -1 ± 5i (one part is real and another part is imaginary)

spiral toward origin

Limit Cycles

A closed path C of a non – linear autonomous system,

dx
= A(x , y)
dt

①
dy
= Q(x , y)
dt

which is approached spirally from either the inside or the outside by a

non closed path C1 of (1) either as t tends to infinity or as t -> ∞ or as t-> -∞
is called limit cycle of (1).

Alternatively, we can say that a limit cycle of the system (1) is an isolated
closed trajectory of (1), isolated means that neighbouring trajectories are not
closed.
 Limit Cycle
Types of Limit cycles
1. Stable limit cycle
if all neighbouring trajectories approach (from inside or outside) the limit
cycle as time approaches infinity then the limit cycle is stable or attracting.

Stable limit cycle

2. Unstable limit cycle

if all neighbouring trajectories are away from a limit cycle , then it is said unstable limit cycle
 Unstable Limit Cycle

3. Semi – stable limit cycle or half stable limit cycle

A limit cycle can be neither stable nor unstable is said to be half stable
limit cycle.

Semi stable limit cycle

Real life Significance limit Cycles-

1. Biology
 Heartbeats and breathing rhythms follow limit cycles, helping maintain stable
biological patterns.
2. Ecology
Predator prey populations rise and fall periodically forming ecological limit cycle.
Limit cycle is non – linearl phenomena , it can not occur in a linear
System.

A linear system,

dx
= ax +by
dt

dy
dt
= cx +dy

can have closed orbits, but they won’t be isolated.

Example 1:-
Show that the following dynamically system possess a limit cycle,
dx
= y + x (1−x 2− y 2 )
dt
①

dy
= −x + y (1−x 2− y 2)
dt

Also study the nature of the trivial equilibrium point.

Solution:-
The corresponding linear system of (1) is given by,
dx
= x+ y
dt

②
dy
= −x + y
dt

( )(
dx
dt
dy
1 1
= −1 1 x
y )( )
dt
¿ ¿
If (x , y )be a critical ( equilibrium or singular ) point of (1) and (2)
( dxdt ) (x , y )
¿ ¿
=0= ( dydt )
( x ¿ , y ¿)

Clearly, (0,0) is a critical point of (1) now the characteristic equation (2) is,
λ2 – (a+d)λ + (ad-bc) = 0
λ2 - 2λ + 2 = 0
λ=1±i (where i2 = -1 )
since, Re (λ) > 0 , so the origin (0,0) is unstable spiral point for both the linear and the non
linear system.
To study the system, we introduce polar coordinates(r,ϴ )
Where, x= r cosϴ
(3)
y= r sin ϴ
Now,
dx dy
+ y =x [ y + x ( 1−x − y ) ] + y ¿ from (1)
2 2
x
dt dt

¿ xy + x 2−x 2 ( x 2 + y 2 )−xy + y 2− y 2 (x 2+ y 2 )

¿ x 2−x 2 ( x 2 + y 2 ) + y 2− y 2 (x 2 + y 2 )
2 2 2 2
¿ x + y (1−x − y )

0r,
dr 2
=r ( 1−r ) where r 2=x 2 + y 2
2
r
dt
dr 2 dr dx dy
=r (1−r ) 2x =2 x +2 y
dt dt dt dt

similarly,
dx dy
−x = y [ y+ x ( 1−x − y ) ]−x [−x+ y ( 1−x − y ) ]
2 2 2 2
y
dt dt
2 2
¿x +y
dx dy
-r2dϴ/dt = r2 where y dt + x dt

dϴ/dt = -1 = r sinϴ( -r sinϴ dϴ/dt) – r cosϴ(r cosϴ dϴ/dt)

= -r2 dϴ/dt (sin 2 ϴ+ cos 2 ϴ)
= -r2 dϴ/dt
Thus, in polar coordinates the system (1) become ,

dr
=r ( 1−r )
2
dt

④
dΘ
=−1
dt

from, dϴ/dt = -1 are get by integrating

∫ d Θ=−∫ dt+t 0 [ where t0 is arbitrary constant ]

Θ=−t+t 0

From,
dr
=r ( 1−r ) integrating by separating variable
2
dt
dr
2
=dt
r (1−r )
1 1 1
+ − dr =dt
r 2 ( 1−r ) 2(1−r 2) ❑
2

ln r 2 −ln |1−r 2|=2 t+ ln|C0|

ln ¿
1
r= The solution of the system (4) may written ,
√ 1−c e−2 t
⑤
Θ=−t+t 0

We may choose t0 = 0 then ϴ = -t ,using (3) the solution of the system (1) becomes
cos t
x=
√1+ c e−2 t
⑥
−sin t
y=
√ 1+c e−2t
The solution (6) and (1) define the path of (1) in the xy plane.
Examining these paths for various values of c ,we note the following conclusion :
a) If c=0 , the path (6) is given by a circle ,
2
x + y =1
2
⑦
Describe in the clockwise direction.

b) If c ≠ 0, the path defines by (6) are not closed paths but rather paths having a special
behaviour.
 If c>0 the path are spiral lying inside the circle (7).
 If c<0 the path lie outside the circle (7).

we conclude that this circle is a stable limit cycle of the system (1).
Example 2 :-
Show that the following dynamically system possess a limit cycle.

dx 2 2
=μx + y + x (x + y )
dt

①
dy ❑ ❑
=−x+ μy +❑ ❑ ¿
dt

(μ is a parameter ). Also study the nature of the trivial

equilibrium point.
Solution :-
The corresponding linear system of the given non- linear system is given by,
dx
=μx + y
dt

②
dy
=−x+ μy
dt

( )(
dy
dt μ 1 x
or the matrix form, dy = −1 μ y )( )
dt

obviously, (0,0) is the trivial equilibrium point of (1), as well as (2) ,

The characteristic equation of (2) ,

( μ−λ
−1
1
μ−λ )
=0 or λ 2−2 μλ+ ( μ2 +1 )=0

λ=μ ±i , ( i 2=−1 )

The eigenvalues are μ+i and μ-I ,

Case I; if μ<0, then (0,0) is asymptotically stable spiral equilibrium point.
Case II: if μ>0 , then (0,0) is an unstable spiral equilibrium point.
To study the system, we get
x=r cos θ , y =r sinθ

Now, r 2=x 2 + y 2
Differentiating both side with respect to ‘t’ ,
dr dx dy
2r =2 x +2 y
dt dt dt
dx
=x ( μx + y + x ( x + y ) ) + y ¿
2 2
r
dt
dr 2
=r (r + μ)
dt
dx dy
Now, y dt −x dt = y ( μx + y + x ( x + y ) )−x (−x + μy + ( x + y ))
2 2 2 2

2 2 2
¿ x + y =r
2 dθ 2
As, −r dt =r

dθ
As, dt =−1 so ,ϴ decrease and clockwise direction.

Thus, the given system become in polar coordinates,

dr 2
=r (r + μ)
dt

(3)
dθ
=−1
dt

Now, ∫ dθ=∫ dt+t 0 [ t0 is the arbitrary constant ]

θ=−t+t 0

dr 2
Now, dt =r (r + μ)

dr
∫ r (r 2+ μ) =∫ dt
1 1
− 2
μ r r +μ[
r
dr=dt
]
On integrating , we get
2 −μ
r= −2 μt where C = 1/C02μ
−C e +1

The solution of the system (3) is given by,

r=
√ −μ
−C e−2 μt +1

④
θ=−t+t 0
Choose t0 = 0 then ϴ = -t, so the solution of the system (1),

x=
√ −μ
−C e−2 μt +1
cos t

y=−
√ −μ
−C e−2 μt +1
sin t

If C=0 the path (5) is given by

2
x + y =μ
2
⑥ (μ<0)
If C<0 (with μ<0 ), the path are spirals lying inside the circle (6)
if C >0 ( with μ< 0 ) the path lie outside the circle.
So, x 2+ y 2=−μ is an unstable limit cycle.
Poincare Bendixson
theorem
System of Equations:
Consider a two-dimensional nonlinear system:

x● = f ( x,y)
y● = g ( x,y)

1) Bounded (Trapping) Region- If there exists a bounded region R in the x,y-plane such
that:
 Any trajectory that starts in R remains inside R forever,
 Then R is called a trapping region (because once a trajectory enters R, it can never
escape).

2) Theorem – If a trajectory starts in R and stays there forever, then the region R must
contain either:
 A stable fixed point, or
 A limit cycle (a closed periodic orbit).

 Properties –
 The theorem can be applied only to systems with two differential equations.
 It applies only to nonlinear systems.
 It describes local stability, not global stability.

 Physical Meaning – The system’s motion is limited it cannot wander indefinitely.

Inside the bounded region, the system’s behavior becomes predictable:
 It either settles at a steady point (fixed point)
 Repeats its motion in a closed path (limit cycle)
 Trajectory in Trapping regin

5) Diagram Explanation –
 R (Orange Boundary) Trapping region, all motion stays inside it
 Green Arrows – Trajectories showing the motion of the system
 P (Red Dot) – Fixed point, equilibrium where system can settle
 Pink Loop – Limit cycle, a closed path showing periodic motion

Reference - Differential Equations and mathematical Biology ( 2nd

edition ) by D. S. Jones Plank, B. D. Sleeman