ASSIGNMENT

ON
Existence and Uniqueness
Theory

COURSE TITLE: Population dynamics

COURSE NO: MATH-5209

SUBMITTED TO SUBMITTED BY
Dr. Md. Haider Ali Biswas Tahera Parvin
Professor Student ID: M.Sc- 201203
MATHEMATICS DISCIPLINE
MATHEMATICS DISCIPLINE KHULNA UNIVERSITY, KHULNA
KHULNA UNIVERSITY, KHULNA

SUBMISSION DATE: 26.09.2020

MATHEMATICS DISCIPLINE, KHULNA UNIVERSITY, KHULNA

Mathematical model
A mathematical model is a description of a system using mathematical concepts and language.
The process of developing a mathematical model is termed mathematical modeling.
Classifications
Mathematical models are usually composed of relationships and variables. Relationships can be
described by operators, such as algebraic operators, functions, differential operators, etc.
Variables are abstractions of system parameters of interest, that can be quantified. Several
classification criteria can be used for mathematical models according to their structure:

 Linear vs. nonlinear: If all the operators in a mathematical model exhibit linearity, the
resulting mathematical model is defined as linear. A model is considered to be nonlinear
otherwise. The definition of linearity and nonlinearity is dependent on context, and linear
models may have nonlinear expressions in them. For example, in a statistical linear model, it
is assumed that a relationship is linear in the parameters, but it may be nonlinear in the
predictor variables. Similarly, a differential equation is said to be linear if it can be written
with linear differential operators, but it can still have nonlinear expressions in it. In
a mathematical programming model, if the objective functions and constraints are
represented entirely by linear equations, then the model is regarded as a linear model. If one
or more of the objective functions or constraints are represented with a nonlinear equation,
then the model is known as a nonlinear model.
Linear structure implies that a problem can be decomposed into simpler parts that can be
treated independently and/or analyzed at a different scale and the results obtained will
remain valid for the initial problem when recomposed and rescaled.
Nonlinearity, even in fairly simple systems, is often associated with phenomena such
as chaos and irreversibility. Although there are exceptions, nonlinear systems and models
tend to be more difficult to study than linear ones. A common approach to nonlinear
problems is linearization, but this can be problematic if one is trying to study aspects such as
irreversibility, which are strongly tied to nonlinearity.
 Static vs. dynamic: A dynamic model accounts for time-dependent changes in the state of
the system, while a static (or steady-state) model calculates the system in equilibrium, and
thus is time-invariant. Dynamic models typically are represented by differential
equations or difference equations.
 Explicit vs. implicit: If all of the input parameters of the overall model are known, and the
output parameters can be calculated by a finite series of computations, the model is said to
be explicit. But sometimes it is the output parameters which are known, and the
corresponding inputs must be solved for by an iterative procedure, such as Newton's
method (if the model is linear) or Broyden's method (if non-linear). In such a case the model
is said to be implicit. For example, a jet engine's physical properties such as turbine and
nozzle throat areas can be explicitly calculated given a design thermodynamic cycle (air and
fuel flow rates, pressures, and temperatures) at a specific flight condition and power setting,
but the engine's operating cycles at other flight conditions and power settings cannot be
explicitly calculated from the constant physical properties.
 Discrete vs. continuous: A discrete model treats objects as discrete, such as the particles in
a molecular model or the states in a statistical model; while a continuous model represents
 the objects in a continuous manner, such as the velocity field of fluid in pipe flows,
temperatures and stresses in a solid, and electric field that applies continuously over the
entire model due to a point charge.
 Deterministic vs. probabilistic (stochastic): A deterministic model is one in which every
set of variable states is uniquely determined by parameters in the model and by sets of
previous states of these variables; therefore, a deterministic model always performs the same
way for a given set of initial conditions. Conversely, in a stochastic model—usually called a
"statistical model"—randomness is present, and variable states are not described by unique
values, but rather by probability distributions.
 Deductive, inductive, or floating: A deductive model is a logical structure based on a
theory. An inductive model arises from empirical findings and generalization from them. The
floating model rests on neither theory nor observation, but is merely the invocation of
expected structure. Application of mathematics in social sciences outside of economics has
been criticized for unfounded models. Application of catastrophe theory in science has been
characterized as a floating model.
 Strategic vs non-strategic Models used in game theory are different in a sense that they
model agents with incompatible incentives, such as competing species or bidders in an
auction. Strategic models assume that players are autonomous decision makers who
rationally choose actions that maximize their objective function. A key challenge of using
strategic models is defining and computing solution concepts such as Nash equilibrium. An
interesting property of strategic models is that they separate reasoning about rules of the
game from reasoning about behavior of the players.

Population dynamics, importance of population dynamics and applications

The term "population dynamics" refers to how the number of individuals in a population changes
over time.
Population dynamics deals with the short-term and long-term changes in the size and age
composition of populations, and the biological and environmental processes influencing those
changes.
Population dynamics is the branch of life sciences that studies the size and age composition
of populations as dynamical systems, and the biological and environmental processes driving
them (such as birth and death rates, and by immigration and emigration).
Example scenarios are ageing populations, population growth

Dynamical system

A dynamical system is a system in which a function describes the time dependence of a point in
a geometrical space. Examples include the mathematical models that describe the swinging of a
clock pendulum, the flow of water in a pipe, and the number of fish each spring time in a lake.
At any given time, a dynamical system has a state given by a tuple of real numbers that can be
represented by a point in an appropriate state space. The evolution rule of the dynamical system
is a function that describes what future states follow from the current state. Often the function
is deterministic, that is, for a given time interval only one future state follows from the current
state. However, some systems are stochastic, in that random events also affect the evolution of
the state variables.

Carrying capacity
The carrying capacity of an environment is the maximum population size of a
biological species that can be sustained in that specific environment, given the
food, habitat, water, and other resources available. In population ecology, carrying capacity is
defined as the environment's maximal load, which is different from the concept of population
equilibrium, which may be far below an environment's carrying capacity. The effect of carrying
capacity on population dynamics may be modelled with a logistic function.

Carrying Capacity
The graph above shows the population (N) of a certain species over time (t). At the carrying
capacity (K), the population stops growing as resources are maxed.
Modelling Population
• Mathematical models are required to understand and predict population behavior.
• Mathematical models are widely used in ecology. Ecosystems tend to be very complex and
governed by many intricate and usually non-linear mechanistic interactions.
• This course will be structured upon a foundation from mathematical models, expanding to the
biological evidence to support and/or reject various factors regulating populations.
• Introduction to the formulation, analysis and application of mathematical models that describe
the dynamics of biological populations.

Coupled system
A coupled system is formed of two differential equations with two dependent variables and an
independent variable.
An example –

where a, b, c and d are given constants, and both y and x are functions of t.

Nonlinear Systems

Nonlinear state models

 We will deal with the stability analysis of nonlinear systems, with emphasis on Lyapunov’s
method.

 Consider a dynamical systems that are modeled by a finite number of coupled first-order
ordinary differential equations as


x  f  t , x, u  ............................ 1

We call (1) the state equation and refer to x as the state and u as the input.

 Consider another equation

y  f  t , x, u  ................................  2 

Equation (2) is called the output.

State-space model

 Equations (1) and (2) are called together the state space model or simply state model.

Unforced State equation

 If the state model (1) is taken as


x  f  t , x  ..................... 3 

(3) where the input is specified as a function of time and state u    t , x  , then (3) is called
unforced state equation.

Implicit and explicit functions

In mathematics,

The explicit function is a function in which the dependent variable has been given ―explicitly‖
in terms of the independent variable. Or it is a function in which the dependent variable is
expressed in terms of some independent variables.

It is denoted by:

y=f(x)

Examples of Explicit functions are:

y=axn+bx where a , n and b are constant.

y=5x3-3

The Implicit function is a function in which the dependent variable has not been given
―explicitly‖ in terms of the independent variable. Or it is a function in which the dependent
variable is not expressed in terms of some independent variables.

It is denoted by:

R(x,y) = 0
Some examples of Implicit Functions are:
x2 + y2 – 1 = 0

Initial value problem

In the field of differential equations, an initial value problem is an ordinary differential

equation together with a specified value, called the initial condition, of the unknown function at a
given point in the domain of the solution.

Boundary value problem

In mathematics, in the field of differential equations, a boundary value problem is a differential

equation together with a set of additional constraints, called the boundary conditions. A solution
to a boundary value problem is a solution to the differential equation which also satisfies the
boundary conditions.

Solution of Differential Equation:

Let us consider an initial value problem


x(t )  f (t, x(t ))
 (1)
 x(t0 )  t0

Let g (t ) be a real function defined  t in a real interval I . The function g (t ) is called the
solution of the differential equation (1) on I if

i ) f (t , g (t )) is defined  t  I and
ii ) f (t , g (t ))  0  t  I

i.e. the substitution of g (t ) and its various derivatives for x and its corresponding derivatives in
equation (1) reduces to an identity on I .

Integral Form of the Solution:

Consider the initial value problem (1), then x (t ) be the solution of (1) if it can be written as,
 t
x(t )  x0   f (s, x(s))ds
t0
(2)

Conversely, By differentiating (2) we get, x(t )  f (t, x(t ))

The unknown function at the different point in the domain of the solution.

Mathematically,


x(t )  f (t, x(t ))

 x(t0 )  a, x(t1 )  b)

With f : D    n  n and (t0 , a), (t1 , a)  D

The basic difference between initial value problem and boundary value problem be if the
independent variable is time over the domain [0,1] , a boundary value problem would specify
value for x (t ) at both t  0 and t  1. Where as an initial value problem would specify a value of x (t )
and x(t ) at time t  0.

For example,

Initial Value Problem:

x(t )  x(t )  0

with initial condition x(0)  1, x(0)  0

Boundary Value Problem:

x(t )  x(t )  0
x(0)  1, x(1)  

Compact Set

A set K in a metric space (X, d) is said to be compact if any open cover U A of K has a finite
n
sub-cover. In other word, a set K is compact in if and only if it is closed and bounded.
Bounded Function:

A function that is not bounded is said to be unbounded. Sometimes, if f(x) ≤ A for all x in X,
then the function is said to be bounded above by A. On the other hand, if f(x) ≥ B for all x in X,
then the function is said to be bounded below by B. The concept should not be confused with
that of a bounded operator.

Continuous and discontinuous function:

Left figure is the continuous function and right figure is the discontinuous function.

In calculus, a continuous function is a real-valued function whose graph does not have any
breaks or holes, otherwise discontinuous function.

Another way of definition,

In mathematics, a continuous function is a function that does not have any abrupt changes
.More precisely, sufficiently small changes in the input of a continuous function result in
arbitrarily small changes in its output. A continuous function with a continuous inverse
function is called a homeomorphism.

Absolutely Continuous function:

Let I be an interval in the real line R . A function f :I R is absolutely

continuous on I if for every positive number  , there is a positive number  such that
whenever a finite sequence of pairwise disjoint sub-intervals ( xk , yk )
of I with ( xk , yk )  I satisfies.

(y
k
k  xk )  
 then k
f ( yk )  f ( xk )  

The collection of all absolutely continuous functions on I is denoted AC ( I ) .

A continuous function fails to be absolutely continuous if it fails to be uniformly continuous,
which can happen if the domain of the function is not compact – examples are tan(x) over
[0, π/2), x2 over the entire real line, and sin(1/x) over (0, 1]

cantor function

In mathematics, the Cantor function is an example of a function that is continuous, but

not absolutely continuous.

Uniformly Continuous Function:

Let f be a function, let M be a subset of its domain D(f). We say that f is uniformly
continuous on M if for every   0 there exists   0 such that for all a,x from M satisfying |x -
a| <  we have | f(x) - f(a)| <  .

Example: Consider the function f(x)=1/x. We know that it is continuous on (0,  ). However, it
is not uniformly continuous on this interval. Indeed, in the picture we tried to show that given a
certain epsilon, we are forced to make delta smaller and smaller to win games for this epsilon as
we move closer to 0 with a and it seems that this process does not stop, no delta will work
universally.

Piecewise continuous function:

A function is called piecewise continuous on an interval if the interval can be broken into a finite
number of subintervals on which the function is continuous on each open subinterval (i.e. The
subinterval without its endpoints) and has a finite limit at the endpoints of each subinterval.

Fig: piecewise continuous function

Lipschitz continuous function:

Lipschitz continuity, named after Rudolf Lipschitz, is a strong form of uniform

continuity for functions. Intuitively, a Lipschitz continuous function is limited in how fast it can
change: there exists a real number such that, for every pair of points on the graph of this
function, the absolute value of the slope of the line connecting them is not greater than this real
number; the smallest such bound is called the Lipschitz constant of the function For instance,
every function that has bounded first derivatives is Lipschitz.

Fig: Lipschitz continuous function

Convex Set:

In mathematics, a real valued function defined on an n-dimensional interval is

called convex (or convex downward or concave upward) if the line segment between any two
points on the graph of the function lies above or on the graph, in a Euclidean Space (or more
generally a vector space) of at least two dimensions. Equivalently, a function is convex if
its epigraph (the set of points on or above the graph of the function) is a convex set.

Convex Function:

Let X be a convex set in a real vector space and let f : X   be a function.

 f is convex if:  x1 , x2  X t  0,1 : f  tx1  (1  t ) x2   t f  x1 

 f is called strictly convex if:

 x1  x2  X , t  0,1 : f  tx1  (1  t ) x2   t f  x1   1  t  f  x2 

Fig: Convex function

 A function (in black) is convex if and only if the region above its graph (in green) is a
convex set.
Convex Function play an important role in many areas in mathematics. They are especially
important in the study of optimization problem Where they are distinguished by a number of
convenient properties.

Properties of Convex Function:

Function of one variable:

1. Suppose f is a function of one real variable defined on an interval, and let

f  x1   f  x2 
R  x1 , x2   . (note that R  x1 , x2  )is the slope of the purple line in the
x1  x2

above drawing: note also that the function R is symmetric in  x1 , x2  . f is convex iff

R  x1 , x2  is monotonically non-decreasing in x1
 2. A convex function f one real variable defined on sum open interval c is continuous on
c and Lipschitz continuous on any closed sub interval.
A differentiable function of one variable is convex on an interval if and only if its derivative is
monotonically non-decreasing on that interval. If a function is differentiable and convex then it is
also continuously differentiable.

Convexity:

Convexity is a measure of the curvature, or the degree of the curve, in the relationship between
bond prices and bond yields. Convexity demonstrates how the duration of a bond changes as the
interest rate changes.

Trajectory

In a dynamic system, a trajectory is the set of points in state space that are the future spaces
resulting from a given state.

Orthogonal Trajectories

Let

F ( x, y, e)  0 (1)

Let a given one parameter family of curses in the xy plane.

A curve that intersects the curves of the family (1) at right angle is called an orthogonal
trajectory of the given family.

For example: consider the family of circles.

x2  y 2  c2 (2)

with Centre at the origin and radius c.

Each straight line through the origin y  kx is an orthogonal trajectory of the family of circle (2).
 y

function:

A function is a function that is differentiable for all degrees of differentiation. For

instance, is because its th derivative exists and is continuous.

All polynomials are . functions are also called "smooth" because neither they nor their
derivatives have "corners," which would make their graph look somewhat rough. For
example, is not smooth.

Differentiable function:
In calculus (a branch of mathematics), a differentiable function of one real variable is a function
whose derivative exists at each point in its domain. As a result, the graph of a differentiable
function must have a (non-vertical) tangent line at each interior point in its domain, be relatively
smooth, and cannot contain any breaks, bends.
A differentiable function

Differentiability and Continuity

If f is differentiable at a point x0, then f must also be continuous at x0. In particular, any
differentiable function must be continuous at every point in its domain. The converse does not
hold: a continuous function need not be differentiable.

The absolute value function is continuous (i.e. it has no gaps). It is differentiable everywhere
except at the point x = 0, where it makes a sharp turn as it crosses the y-axis.

Smooth function:
A smooth function is a function that has continuous derivatives up to some desired order over
3
some domain.Now consider g(x)= x , this other function is also smooth.

Here is the graph of g(x):

 dg x 
Here is graph of  3x 2 :
dx

d 2 g x 
Here is graph of  6x
dx 2

Non-Smooth function:
h(x)=|x| is not smooth, because it has corner.
Here is how the derivative of h(x) looks like:

Here is the derivate of h(x) as you can see its derivative is not even continuous.

Autonomous system

In mathematics, an autonomous system or autonomous differential equation is

a system of ordinary differential equations which does not explicitly depend on the independent
variable. When the variable is time, they are also called time-invariant systems.
An autonomous system is a system of ordinary differential equations of the form

x  f  x

If the system is not autonomous, then it is called non-autonomous or time varying.

Lemma 1:

Let x  ( x) be a function with continuous derivative, define in the interval

I h ( x0 )  [ x0  h, x0  h], with values in [y0  b, y0  b].

Then  satisfies the initial value problem

  ' ( x)  F ( x, ( x))
(1) 
( x0 )  y 0
If and only if it satisfy the integral equation
x
(2) ( x)  y0   F (t , (t )dt.
x0

Proof: Let us first assume that  is differentiable so that  satisfies (1). Then we integrate and
deduce that for x  x0  h

x x
x0
' (t )dt  F (t , (t )dt.
x0

However the left hand side is equal to ( x)  ( x0 ) by the fundamental theorem of calculus.

Thus, since ( x0 )  y0 we get

x
( x)  y0   F (t , (t )dt.
x0

Which is equivalent to (2) .

Vice versa, assume that  is continuous and satisfies the integral equation (2). Then the
integrand F (t , (t )) is also a continuous function of t and thus, by the fundamental theorem of
x
calculus the integral 
x0
F (t , (t )dt. is a differentiable function of x with derivative F (x, (x)

Thus the right hand side of (2) is a differentiable function with derivative F (x, (x) and by (2)

 is differentiable with  ' ( x)  F (x, (x) which is one part of (1). It remains to check that
x0
( x0 )  y0 but this immediately follows from (2), since,  dt  0 .
x0

Thus the equation (2) has a unique solution in [ x0  h, x0  h]

Lemma 2:

The following is true:

If g satisfy g ( x)  y0  b for all x in x0  h, x0  h then Tg also satisfies Tg ( x)  y0  b for all x in

x0  h, x0  h

Proof:
It is here where the crucial h  M is used. If g ( x)  y0  b and x  x0  h  min(a, m) then
b b

for x0  x

Tg ( x)  y0 
 F t, g (t)  dt
x0


 F t, g (t) dt
x0
x


 Mdt  M x  x
x0
0  Mh  b

x x

And a similar argument goes for x  x0 (just write  ....... dt instead of  ....... dt in the case)
x0 x0

Now for the iteration alluded to above we set,

0 ( x)  y0
x
1 ( x)  T0 ( x)  y0   F (t , y0 )dt
x0
x
2 ( x)  T1 ( x)  y0   F (t ,1 (t ))dt
x0

x
2 ( x)  T 1 ( x)  y0   F (t ,1 (t ))dt
x0
x
3 ( x)  T 2 ( x)  y0   F (t ,2 (t ))dt
x0

If n 1 is already defined, we set

x
n ( x)  y0   F (t ,n1 (t ))dt
x0
Lemma 2 says that if ( x, n1 ( x)) belongs to the rectangle R where F was defined and if
t  x0  h . Thus the definition makes indeed sense for all n provided.

We repeat: by using lemma 2 we can define the sequence n ( x) recursively by,

0 ( x)  y0

and

n ( x)  Tn1 ( x), n  1, 2,............

if x  x0  h

We want the sequence n ( x) to converge to the solution of the initial value problem.

Lemma 3:

Let M 0  max|t  x0 |h | F  t , y0  | .

| x  x0 |n1
The sequence n testifies the estimate | n1  x   n  x  | M 0 K n for all [ x0  h, x0  h]
 n  1!
The sequence n converges to a limit  which is a solution of the initial value problem as well as
the equivalent integral equation.

Proof:

We shall prove the inequality for x0  x  x0  h , and after simple modifications the argument
also yields the desired estimate for x0  h  x  x0 .

We show the inequality by iteration i.e. ―mathematical induction‖. That is we first prove the
estimate for n=0. Then we show that for all n the estimate for n  x   n 1  x  implies the
estimate for n 1  n .

This means because we have shown it for n=0, it then follows for n=1. Since the estimate for n=1
implies the estimate for n=2 the estimate for n=2 is also true. Since the estimate for n=2 implies
the estimate for n=3 the estimate for n=3 is also true, and so on.
First step, we show the assertion for n=0, that is show that

1  x   0  x   M 0 x  x0 if x  x0  h

Indeed 1  T 0 and therefore

x
1  x   0  x    F t , y  dt
0
x0

x
  F t , y  dt
x0
0

 M 0 x  x0

Second step, Now we show that for all n the inequality

x  x0
n

n1  n  x   n 1  x   M 0 k n 1

n!

Implies the inequality

n 1
x  x0
n  n 1  x   n  x   Mk n

 n  1!
We write

n 1  x   n  x   Tn  x   T n 1  x 
x
  F t , t   F t , t  dt
x0
n n 1

By the mean value theorem of differential calculus with respect to the y variable

F
F  t , n  t    F  t , n 1  t    t ,   n t   n1 t 
y

Where  is some value between n  t  and n 1  t  . Therefore

F  t , n  t    F  t , n 1  t    k n  t   n 1  t 

But since we assume the validity of n 1  we have

 t  x0
n

n  t   n 1  t   M 0 k n 1

n!

Thus we get

 F t ,  t   F t ,  t  dt
x0
n n 1

x
  F t ,  t   F t ,  t  dt
x0
n n 1

 t  x0 
x n

 k  M 0k n 1
dt
x0
n!

 t  x0 
x n

 M 0k n 1
k dt
x0
n!

 x  x0 
n

 M 0k n

 n  1!

Putting inequalities together we get n  which we wanted to verify

Theorem 1: (Existence theorem):

Let us consider an IVP

.
y ( x )  f  x, y 

y  x0   y0 …………………….(1)

Suppose that f  x, y  is continuous function in some region *( ): | |

| | +( 0).

Since f is continuous in a closed and bounded domain, it is necessary bounded in R, i.e there
exists k 0 such that,

| ( )| ( ) .

Then the IVP (1) has at least one solution ( ) define in the interval | |
 Where, { ).

Proof: Let us consider a solution in the IVP (1) which is equation of the integral equation,

( ) ∫ ( ( )) . . (2)

On the interval | |

| | ( )

From the picard’s iteration method, we get

( )

( ) ∫ , ( )-

( ) ∫ , ( )- .2( )

Where

n=1,2,3…

We shall first show that the sequence { ( )+ are constant and lies in the rectangle | | .

We note that, ( ) is continuous because, constant function is always continuous.

Now,

( ) ∫ , ( )-

Since , ( )- ( )

Similarly, we can show that,

 ( ) ∫ , ( )- .…………………………..(3)

Are constant for 2 ..

Clearly from (3), | ( ) | |∫ ( ( )|

|∫ ( ( )|

∫ | |

| |

So,{ ( )+ are constant and lies within R.

For | | .

Next we show that the sequence * ( )+ converges uniformly to a constant function on

| | .

Let us consider the infinite series , ( ) ( )- , ( ) ( )-

, ( ) ( )- . ..( )

Now (1) | ( ) ( )| |∫ ( ( )

| | .

(2) | ( ) ( )| |∫ * , ( )- , ( )+ |

∫ | ( ) ( )|

∫ | |
 | |

Similarly,

| ( ) ( )| |∫ * , ( )- * ( )-+ |

| ( ) ( )| (5)

Or, ∑ | ( ) ( )| ∑

( )
∑ ..............................................................(6)

From weierstrass M-test, we have if { ( )+ be a sequence of function with,

A 
ym  x   M m it in  a, b  Write M n <0 then  y  x  converges
n uniformly in  a, b  to a
m  m 

unique function y(x).


Applying weierstrass M-test in (6) ,if follow that y0  x     yn  x   ym1  x  converges
m t

absolutely and uniformly on x  x0  h and hence its converges to a function in this interval i.e
yn  x   y  x 

Then yn(x) converges uniformly to a continuous function in the interval x  x0  h . Now we

prove that y(x) in a solution of (1) on x  x0  h .

Since f(x,y) is cont. is relation of y and yn  x   y  x  as n  

We have it lim itn  x, yn  x    f [ x, y ( x)] (7)

y(x)= lim it y  yn  x 

x
= y0  lim itn  f t , yn  t dt [by 1a]
x0

x
= y0   lim itn f t , yn  t   dt [by -7]
x0

x
 y  x   y0   f  t , y  t  dt
x0

Hence y(x) is a continuous solution of (1)

Theorem 2(Uniqueness theorem):

Let us consider an IVP

.
x( ) ( )………………..(1)

( )

Suppose that ( ) is continuous in some region *( ): | | | | +

( 0).

Since f is continuous in a closed and bounded domain, it is necessary bounded in R, i.e there
exists k 0 such that,

| ( )| ( ) .

Then the IVP (1) has at least one solution ( ) define in the interval | |

Where, { ).

Lipchitz condition give that f t , x1   f t , x 2   L x1  x 2 t , x  R (1a)

Lastly we show that x(t) is the solution of (1) . Suppose that Z(t) is also a continuous solution of
(1) for t  t 0  h
Then ,
t
z  t   x0   f  s, x  s  ds
t0

Let N=max xt   z t  …………………………………………………………………. (8)

Then for t  t 0  h we obtain

t
x  t   z  t    f  s, x0  s   f  s, z  s  ds
t0

t
 L  x0 s   z s ds [ by 1(a)and 8]
t0

 LN t  t 0  LNh ……………………………………………………………………….... (9)

And
t
x2  t   z  t    f  s, x1  s    f  s, z  s   ds
t0
t
 L  x1  s   z  s  ds
t0
t
 L N  s  s0 ds[by 9]
2

t0

L2 Nh 2

2 !

In General,
NLn h n
xn (t )  z (t ) 
n!
If n   then,
( Lh) n
xn (t )  z (t )  N lim (10)
n  n!
Since the right side of (10) approaches to zero. It follows that x(t )  z (t ) .

This implies that x(t ) and z (t ) are identical.

So, x  t  is the only solution of (1).

Hence, the initial value problem.(1) has a unique solution x  x(t ) on the interval x  x0  a .

Theorem 3: State and prove the existence and uniqueness theorem depending on the arbitrary
function.

Statement: If in a domain D , if xy plane

1. (i) f ( x, y) is continuous and satisfies a Lipschitz condition with respect to y ,with

Lipschitz constant k.

(ii) F ( x, y ) is constant in D

(iii) F ( x, y )  f ( x, y )   for ( x, y )  D
2. ( x0 , y0 ) is a point if D and
(i)  ( x) is the solution of IVP
dy
 f ( x, y ), y ( x0 )  y0
dx
(ii) ( x) is the solution of
dy
 F ( x, y ), y ( x0 )  y0
dx
(iii)  x,  ( x)  and  x, ( x)  D thus

 ( x)   ( x)  (e kh  1) on x  x0  h
k

Proof:

We have,

dy
 f ( x. y ), y ( x0 )  y0 (1)
dx

dy
&  F ( x, y ), y ( x0 )  y0 (2)
dx

Where, F ( x, y )  f ( x, y )   (3)

We now, assume that, F ( x, y)  f ( x, y)   ( x, y) (4)

Thus,
 ( x, y )  F ( x, y )  f ( x , y )
  ( x, y )   by (3)
Again,

Let,  0 ( x)   ( x) (4a)

x
When,  n ( x)  y0   f t ,  n 1 (t ) dt , n  1, 2... (5)
x0

Then,  ( x)  lim  n ( x) is a solution of (1)

n 

By hypothesis 1(i), the IVP (1) has a unique solution on x  x0  h ,

Again from 2(ii), we have

x
 ( x)  y0   F t , (t1 ) dt on x  x0  h. (6)
x0

Now,

x x
 1 ( x)   ( x)  y0   f t ,  0 (t )  dt  y0   F t , (t )  dt
x0 x0

x x x
  f t ,  0 (t )  dt   F t , (t )  dt
x0 x0 x0

x x x
  f t , (t )  dt   F t , (t )  dt [By (4b)]
x0 x0 x0

  1 ( x)   ( x)  k  ( x  x0 )

Similarly,
 x x
 2 ( x)   ( x)  y0   f t ,  1 (t )  dt  y0   F t , (t )  dt
x0 x0

 k ( x  x0 ) 2 
  ( x  x0 )  
 2! 

2
k ( x  x0 )
j 1 j
 
j 0 j!

Continuity this process and taking limit, we get

n
k j 1 ( x  x0 ) j
lim  n ( x)   ( x)  lim  
n  n  j!
j 1

 (kh) j n
 lim 
k n j 1 j !


  ( x)   ( x) 
k
e kh
 1


Thus,  ( x)   ( x) 
k
e kh
 1 [proved]

Local Existence and Uniqueness Theorem

Statement: Let f (t , x) be piecewise continuous in t and satisfy the Lipschitz condition
 
f (t , x)  f (t , y )  L x  y x, y  B  x  R n x  x0  r , t  t0 , t1  . Then, there exists some
  0 such that the state equation x  f (t , x) with x(t0 )  x0 has a unique solution over
 t0 , t0    .

Proof: If x(t ) is a solution of

x  f (t , x), x(t0 )  x0 (1)

Then, by integration, we have

t
x(t )  x0   f ( s, x( s ))ds (2)
t0

Conversely, if x(t) satisfies (2) then x(t) satisfies (1). Thus, the study of existence and uniqueness
of the solution of the differential equation (1) is equivalent to the study of existence and
uniqueness of the solution of the integral equation (2). Considering the right-side of (2) as a
mapping of the continuous function x : t0 , t1   R n , and denoting it by  Px  (t ) we can re-write (2)
as

x(t )   Px  (t ) (3)

 Px  (t ) is continuous in t. A solution of (3) is a fixed point of the mapping P that maps x into Px .
Existence of a fixed point of (3) can be established by using the contraction mapping theorem.
This requires defining a Banach space X and a closed set S  X such that P maps S into S and is
contradiction over S .

Let,

X  C[t0 , t0   ], with norm x  max x(t )

C t[ t0 ,t0  ]

and

S  x  X : x  x0 C
 r

Where r is the radius of the ball B and  is a positive constant to be chose. We will restrict the
choice of  to satisfy   t1  t0 so that [t0 , t0   ]  [t0 , t1 ] . By definition, P maps X into X . To
show that it maps S into S ,we can write
t

 Px  (t )  x0  t f  s, x(s)  ds    f  s, x(s)   f  s, x0   f  s, x0 ds

t

0
t0

Bounded on [t0 , t1 ] .Let

h  max f  t , x0 
t[ t0 ,t1 ]

Using the Lipschitz condition (3.2) and the fact that for each x  S ,

x(t )  x0  r , t  [to , t0   ]

We obtain
t

 Px  (t )  x0    f  s, x( s )   f  s, x0   f  s, x0  ds
t0
 t t
   L x( s)  x0  h  ds   ( Lr  h)ds
t0 t0

 (t  t0 )( Lr  h)   ( Lr  h)

And

Px  x0 C
 max
t[ t0 ,t0  ]
 Px  (t )  x0   ( Lr  h)

Hence, choosing   r / ( Lr  h) ensures that P maps S into S . To show that P is a contraction

mapping over S , let x, y  S and consider

t
( Px)(t )  ( Py )(t )    f  s, x(s)   f  s, y(s)  ds 
t0

t
  f  s, x( s )   f  s, y ( s )  ds
t0

t t
  L x( s)  y ( s) ds   dsL x  y
t0 t0 C

Therefore,


Px  Py C
 L x  y C
  x y C
for  
L

Thus, choosing   1 and    / L ensures that P is a contraction mapping over S . By the

contraction mapping theorem, we can conclude that if  is chosen to satisfy

 r 
  min t1  t0 , ,  for   1
 Lr  h L 

Then (2) will have a unique solution in S . This is not the end of the proof, though, since we are
interested in establishing uniqueness of the solution among all continuous functions x(t ) ,that is,
uniqueness in X . It turns out that any solution of (2) in X will lie in S . Since x  t0   x0 is inside
the ball B , any continuous solution x(t ) must lie inside B for some interval of time. Suppose that
x(t ) leaves the ball B . Then,

x(t0   )  x0  r
On the other hand, for all t  t0   ,

x(t )  x0    f  s, x( s )   f ( s, x0 )  f ( s, x0 )  ds

t t
   L x(s)  x0  h  ds   ( Lr  h)ds
t0 t0

Therefore,

r
r  x(t0   )  x0  ( Lr  h)     
Lr  h

Hence, the solution x(t ) cannot leave the set B within the time interval [t0 , t0   ] , which implies
that any solution in X lies in S . Consequently, uniqueness of the solution in S implies uniqueness
in X .

Now we give some examples to examine the existence and uniqueness of differential
equation:

Example 1:

We consider the problem,

e y ( x ) 1
2

y( x) 
(1) 1  x 2 y ( x) 2
y (2)  1

Which we likely cannot solve explicitly.

We want to find an interval on which a solution surely exists. Here our function F is defined by
F ( x, y )  e y 1 (1  x 2 y 2 ) 1 ( x) 2 and x0 , y0 are given by x0  2, y0  1 . Thus we need to pick a
2

rectangle R which is centered at (-2, 1). In this rectangle we need to have good control on F and
F / y and so we certainly have to choose R so small that it contains no points at which the
denominator 1  x 2 y 2 vanished. The exact choice of the rectangle is up to you but the properties of
F and F / y , as required in the theorem, must be satisfied.

Let’s pick a, b small in the definition of R, say let’s choose a=1/2 and b=1/4 so that we work in
the rectangle,

R  {( x, y) : 5 / 2  x  3 / 2,3 / 4  y  5 / 4}.
Notice, that then for ( x, y) in R we have x 2  9 / 4, y 2  9 /16, and therefore, x 2 y 2  81/ 64, so
1
|1  x 2 y 2 | 17 / 64  1/ 4. for ( x, y) in R. Thus we get | (1  x 2 y 2 )1 | 4 and e y  e9/16  3 which
2

implies,

e y 1
2

| F ( x, y ) |  3.4  12 for ( x, y ) in R
1  x2 y 2

Thus a legitimate (but non-optimal) choice for M is =12.

We compute

F 2 ye y 1 e y 1
2 2

( x, y )   2 yx 2
y 1  x 2 y 2 (1  x 2 y 2 )2

Observe that | 2 y | 5 / 2 and | 2 yx 2 | (5 / 2)3 in R and using the bounds above we can estimate for
all (x, y) in R

F 2 ye y 1 e y 1
2 2

( x, y )   2 yx 2
y 1 x y
2 2
(1  x y )
2 2 2

5
 .12  3.42.(5 / 2)3  780
2

Thus we see that K=780

Now if we take

 b
h  min a,   min{1/ 2, (1/ 4) /12}  1/ 48,
 M

Then by the theorem we can be sure that the problem has exactly one solution in the interval [-2-
h, -2+h]. So for example if we choose h=0.02 (which is less than 1/48), we would deduce that
there is a unique solution in the interval [-0.02, -1.98].

Example 2:
Here, y '  xy  sin y , y (0)  2
So, we can write
f ( x, y)  xy  sin y
Define a region
   
i.e. R : ( x, y ) : x  10, y  , 
 2 
f
 x  cos y
y
Which is also continuous.
And here,
f ( x, y )  xy  sin y
 xy  sin y
 
 10.  sin
2 2
 5  1
 k; where, k  5  1

f
 x  cos y
y
 x  cos y
 10  0
 10

f
So, here f and are both continuous and bounded in a closed rectangle about x0  0 and
y
y0  2 .

Hence, there exists unique solution in the neighborhood of  0, 2 

Example-3:
Here given ordinary differential equation are
y '  1  y 2 ; y (0)  0
So, we can write
f ( x, y )  1  y 2
f
  2y
y
Consider the rectangle,
S  ( x, y ) : x  100, y  1

f
Clearly, f and are both continuous in S .
y

Hence, there exists unique solution in the neighborhood of  0, 0 

Now,
f  1  y2
And
f ( x, y )  1  y 2
 1  y2
 1  12
2
Now,
 1
  min 100, 
 2
1

2
1
Hence the theorem guarantee existence of unique solution in x  , which is much smaller than
2
the original interval x  100

Since, the above equation is separable, we can solve it exactly.

Now,
y '  1 y2
dy
  1 y2
dx
dy
  dx
1 y2
Integrating both sides,

tan 1 y  x  c (1)
Using initial condition, we get,
tan 1 0  0  c
c  0
From equation (1) we get,
tan 1 y  x
tan(tan 1 y )  tan x
 y  tan x

  
This solution is valid only in   ,  which is also much smaller than  100,100 but
 2 2
nevertheless bigger than that predicted by the existence and uniqueness theorem.

Example 4:
The given IVP is
y '  x y , y (1)  0
 f ( x, y )  x y

Now,
f ( x, y1 )  f ( x, y2 )  x y1  x y2
 xy1  xy2
 x  y1  y2
 k  y1  y2

Since, f is Lipschitz continuous in the neighborhood of the (1,0)

Hence, it has unique solution around x  1.

Example 5:
The given IVP is
1
y  y  x, y (1)  0
' 3

So we can write,
1
f ( x, y )  y  x
3
Now,
1 1
f ( x, y1 )  f ( x, y2 )  y13  x  y2 3  x

1 1
 y13  y2 3

y1  y2
 2 1 1 2
y1 3  y13 y2 3  y2 3

We take, y2  0 then,

y1  0
f ( x, y1 )  f ( x, 0)  2
y1 3

1
Now, we can take y1 very closed to zero. Then 2
becomes unbounded. Hence the relation,
3
y1

f ( x, y1 )  f ( x, 0)  2 y1  0 does not hold always around a region about (1,0) .

Since, f does not satisfy Lipschitz condition, we cannot say whether unique solution exists or
does not exists.
1
On the other hand y  y  x, y (1)  1 has unique solution around 1,1
' 3

Example 6:
The given IVP is
2y
y'  , y ( x0 )  y0
x
So, we can write,
2y
f ( x, y ) 
x
f 2
And, 
y x
Both are discontinuous at x0  0

Here unique solution exists in an interval about x0 for x0  0

Again, we can solve it manually,
Here,
2y
y' 
x
dy 2 y
 
dx x
dy dx
 2
y x
Integrating both sides we get,
ln y  2 ln x  ln c
 ln y  ln x 2 c
 y  cx 2
Here, it is the general solution.
When, x0  0, there exists no solution with y0  0 . When y0  0, then we have infinite number

2y
of solutions y   x 2 (  any real number) that satisfy the IVP y '  , y (0)  0
x

Theorem-3: Suppose S is either a rectangle

x  x0  a, y  y0  b, (a, b 0),
Or a strip

x  x0  a, y  , (a, 0),
And that f is a real-valued function defined on S such that f exists, is continuous on S, and
y

f
( x, y )  K , (( x, y ) in S ),
y
For some K 0 then f satisfies a Lapschitz condition on S with Lapschitz constant K.

Proof: we have
 v1 f
f ( x, y1 )  f ( x, y2 )   ( x, t )dt ,
v2 y
And hence

v1 f
f ( x, y1 )  f ( x, y2 )  v2 y
( x, t ) dt  K y1  y2 ,

For all ( x, y1 )  ( x, y2 ) in S.

An example of a function satisfying a Lapschitz conditions is

f ( x, y)  xy 2
On

R: x  1, y  1
Here

f
( x, y )  2 xy  2
y
For all ( x, y) on R. This function does not satisfy a Lapschitz condition on the strip

S: x  1, y  ,
Since

f ( x, y1 )  f ( x, 0)
 x y1 ,
y1  0
Which tends to infinity as y1  , if x  0.

An example of a continuous function not satisfying a Lapschitz condition on a rectangle is

2
f ( x, y )  y 3

On

R: x  1, y  1.
Indeed, if y1 0,

2
f ( x, y1 )  f ( x, 0) y 3 1
 1  1 ,
y1  0 y1 y1 3
Which is unbounded as y1  0.
 Picard iteration
Describe the Picard iteration technique for approximate solution for initial value problem.

The successive approximations method also called the Picard iteration method. Provides a
scheme that can be used for solving initial value problem or integral equations. This method
solves any problem by finding successive approximations to the solution by starting with an
initial guess, called the zeroth approximation. As will be seen, the zeroth approximation is any
selective real-valued function that will be used in a recurrence relation to determine the other
approximations.

This method gives approximate solution to the initial value problem. Note that the initial value
problem is equivalent to the integral equation

A rough approximation to the solution y(x) is given by the function y0  x   y0 which is simply a
horizontal line through  x0 , y0  (don’t confuse function y0  x  with constant y0 ). We insert this
to the R.H.S of (1) in order to obtain a better approximate solution, say y1  x  . Thus,

x
y1  x   y0   f t , y0 t  dt
x0

x
 y0   f  t , y0 dt
x0

The next step is to use this y1  x  to generate another approximate solution y2  x  :

x
y2  x   y0   f t , y1 t  dt
x0

At the n-th stage we find

x
yn  x   y0   f  t , yn 1  t  dt
x0

Theorem: If the function f  x, y  satisfy the existence and uniqueness theorem for IVP
y '  f  x, y  ; y  x0   y0 , then the successive approximation yn  x  converges to the unique
solution y(x) of the IVP.
Proof: The given differential equation is equivalent to the integral equation
x
y  x   y0   f t , y t  dt................................. 1
x0

This suggests the iteration procedure (Picard iteration) given by

y0  x   y0 , which is simply a horizontal line through  x0 , y0  .

We insert this to the R.H.S of (1) in order to obtain a better approximate solution, say y1  x  ,
Thus
x
y1  x   y0   f t , y0 t  dt
x0

x
 y0   f  t , y0 dt
x0

The next step is to use this y1  x  to generate another approximate solution y2  x  :

x
y2  x   y0   f t , y1 t  dt
x0

At the n-th stage we find

x
yn  x   y0   f  t , yn 1  t  dt.....................  2 
x0

We show that yn  x  converge uniformly to a solution y(x).

First of all,
x
y1  x   y0  x    f  t , y0  t  dt
0

 Mx

r
This is where we need the hypothesis x  i.e. this condition is necessary to keep our iterates
M
inside B. Here, B be the Ball of radius r centered of  x0 , y0 

Next we notice
 x
y2  x   y1  x    f  t , y1  t    f  t , y0  t   dt
0
x
  L y1  t   y0  t  dt
0
x
x2
  LMtdt  LM
0
2

In general,

M  Lx 
n 1

yn 1  x   yn  x  
L  n  1!

m 1
Since, ym  x   yn  t     yi 1  x   yi  x  
i 1

It follows that,

 Lc 
i 1
m 1
ym  x   y n  x   
i 1  i  1 !

This the sequence yn  x  is uniformly Cauchy by comparison to the power series for eLc .

By the dominated convergence theorem it follows from equation (2) that the limit y(x) is a
solution of the given initial value problem.

Example 1: Apply Picard iteration for the initial value problem

y '  2 x 1  y  , y  0   2

Solution: Here, y0  x   2

We know from the Picard iteration

x
yn  x   y0   f  t , yn 1  t  dt
x0

Now,
 x
y1  x   y0   f t , y0 t  dt
x0

x
y1  x   2   2t 1  2 dt
0
x
 2    2t  4t dt
0
x

 2  t
2
  2t 2
 0

 2  x  2x
2 2

 2  x2
x
y2  x   2   2t t 2  1 dt  
0
x


 2   2t 3  2t dt 
0
x
t4 2 
 2   t 
2  0
4
x
 2  x2
2
x4
 2  x2 
2

 2 t4 
x
y3  x   2   2t  t   1dt
0  2 
x


 2   2t 3  t 5  2t dt 
0
x
 t t t  4 6 2
 2  2  2 
 4 6 2 0
4 6
x x
 2   x2
2 6
x 4 x6
 2  x2  
2 6
 x
 t6 t4 
y4  x   2   2t    t 2  1dt
0 6 2 
x
 t7 5 
 2     t  2t 3  2t dt
0 
3
x
 t8 t 6 t 4 2 
 2   t 
 24 6 2  0
4 6 8
x x x
 2  x2   
2 3! 4!

By induction, it can show that

x4 x6 n x
2n
yn  x   2  x 2    .........   1
2 3! n!
4 6 2n
x x n x
 1  1  x 2    ..........   1
2 3! n!
 1  e x
2

Hence, yn  x   1  e  x as n   . Now y  x   1  e  x is the exact solution of the given initial

2 2

value problem.

Because, here

dy
 2 x(1  y ); y (0)  2.................(1)
dx

dy
  2 xdx
1 y

By integrating we get,

 ln(1  y)  x 2  c

Using initial condition we get c  ln(1)

So we get,
 ln(1  y )  x 2  ln(1)
(1  y ) 1
 ln  x2
1
1
  ex
2

1  y
 1  y  e  x
2

 y  1  e x
2

Thus the Picard iterates converge to the unique solution of the given initial value problem.

Example 2: Solve the following equation y '  xy, y  0   1 using Picard iteration.

Solution: Here, y0  x   1

We know from the Picard iteration

x
yn  x   y0   f  t , yn 1  t  dt
x0

Now,

x
y1  x   y0   f t , y0 t  dt
x0

x
y1  x   1   tdt
0
x
t2 
 1  
 
2 0
2
x
 1
2
 x
 t2 
y2  x   1   t 1  dt
0 
2
 t3 
x
 1    t  dt
0
2
x
t2 t4 
 1   
 2 8  0
2 4
x x
 1 
2 8

x
 t2 t4 
y3  x   1   t 1   dt
0 
2 8
 t3 t5 
x
 1    t   dt
0
2 8
x
t2 t4 t6 
 1    
 2 8 48  0
2 4 6
x x x
 1  
2 8 48
.
.
.

x2 x4 x6
yn  x   1     ....................
2 8 48
x2
e 2

x2 x2
Hence, yn  x   e as n   . Now y  x   e is the exact solution of the given initial value
2 2

problem.

Because, Here
dy
 xy; y  0   1........................ 1
dx

dy
 xdx
y

By integrating,

x2
ln y  c
2

Using initial condition we get, c=0

So,

x2
ln y 
2
x2
y e 2

Thus the Picard iterates converge to the unique solution of the given initial value problem.