Scribd
Upload
63 views11 pages

MA417 Lecture 2

IIT Bombay

Uploaded by

Jittu Yadav
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
63 views11 pages

MA417 Lecture 2

IIT Bombay

Uploaded by

Jittu Yadav
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 11

MA-417-ODE

Lecture 2

Debanjana Mitra

Department of Mathematics
Indian Institute of Technology Bombay
Powai, Mumbai - 76

July 30, 2021


First order ODEs

Method of solving ODE


Linear ODE
First order ODEs
Let y be defined from I to R, where I is a subset of R or I = R, i.e.,
t 7→ y (t) ∈ R, ∀ t ∈ I . Here t is the independent variable and y (·) is
the dependent variable.
Notation. dtd
y (t) will be denoted by y 0 (t). Similarly for higher order
derivatives.

Consider a first order ODE: F (t, y (t), y 0 (t)) = 0, where the coefficient of
y 0 is always non-zero.

Linear first order ODE: If F is linear with respect to the dependent


variable and its derivative, the the ODE is called linear. For example.
I y 0 (t) = a(t)y (t) + b(t), where a(·) and b(·) are continuous
functions defined from R to R.
I t 2 y 0 (t) + t 3 y (t) = 5.

Non-linear first order ODE: If F is not linear with respect to the


dependent variable and its derivative, then the ODE is called non-linear.
For example.
I y 0 (t) = y 2 (t).

I y 0 (t) = t sin(y (t)).


Remark. Let us observe that no real valued solution exists for the ODE

(y 0 (t))2 + y 2 (t) + 1 = 0.

We shall restrict our attention to the first order ODE in the form:

y 0 (t) = f (t, y (t)),

where f : D → R, is a function defined on D, a subset of R2 .


Geometric interpretation: For a given f : D → R, where D ⊂ R2 , solving
y 0 (t) = f (t, y (t)), means finding a differentiable curve y : I → R, where
I ⊂ R, whose graph (i.e., the set of all points (t, y (t)) for all t ∈ I ) has
the slope f (t, y (t)) for each t ∈ I .
Note that the above problem may have many solutions on an interval I .
For example, if the function f ≡ 1 on D, then the ODE: y 0 (t) = 1, has
infinitely many solutions: yc (t) = t + c for any c ∈ R. However, if we
restrict the graph of the solution passing through a prescribed point
(1, 1) on t − y plane, then there exists only one solution of the ODE:
y 0 (t) = 1 with y (1) = 1 and the solution is y (t) = t. This motivates to
study the ODE with an Initial value.
Initial value problem-ODE

Initial value problem (IVP): Consider an ODE with initial value

y 0 (t) = f (t, y (t)), y (t0 ) = y0 ∈ R, (1)

where f : D → R, where D is a subset of R2 and (t0 , y0 ) ∈ D is


prescribed.
Definition (Definition of a solution of IVP-ODE (1))
If there exist an interval I in R containing t0 and a differentiable function
y : I → R such that
1. y (t0 ) = y0 and (t, y (t)) ∈ D for all t ∈ I ,
2. y 0 (t) = f (t, y (t)) for all t ∈ I ,
then y (·) is called a solution of IVP-ODE (1).
Remark. In addition, if f in (1) is continuous on D and there exists a
y (·) solution to (1), then y ∈ C 1 (I ).
Recall the definition of C 1 (I ) (next slide).
Notation: C 1 (I ) is the collection of all differentiable functions on I with
continuous first derivative.
1
Qn. Give an example
 2 where a function is differentiable but not in C .
x sin(1/x), x 6= 0
Ans. g (x) =
0, x = 0.

Ex. Can we solve an ODE: y 0 (t) = h(t) with y (0) = 1, where



−1 t < 0,
h(t) =
1 0 ≤ t.

Ans. No, Darboux’s theorem (Intermidate value property for y 0 (·)).


Method of solving ODE

Here we discuss few methods to solve ODEs explicitly, if possible.


Linear ODE
Let us consider the linear ODE:

y 0 (t) + a(t)y (t) = b(t), y (t0 ) = y0 ,

where a(·) and b(·) are continuous functions defined from (α, β) to R
and t0 ∈ (α, β).

Integration and differentiation: Let g be a Riemann integrable function


on [α, β], For α ≤ x ≤ β, set
Z x
G (x) = g (s) ds.
α

Then G (·) is a continuous function defined on [α, β]. In addition, if


g ∈ C ([α, β]), continuous on [α, β], then G (·) is differentiable on [α, β]
and
G 0 (x) = g (x), ∀ x ∈ [α, β].

Anti-derivative: Let g : I → R be defined. An anti-derivative of g is a


differentiable function G : I → R such that G 0 (x) = g (x), for all x ∈ I .
Example. Let g (·) be a continuous function on I containing a point t0
Rt
and G (t) = t0 g (s) ds. Then G (t0 ) = 0 and G 0 (t) = g (t). Let us set
µ(t) = e G (t) , for all t ∈ I . Then

µ0 (t) = G 0 (t)e G (t) = g (t)µ(t), ∀t ∈ I.

Let us go back to the problem: Solve the linear ODE:

y 0 (t) + a(t)y (t) = b(t), y (t0 ) = y0 ,

where a(·) and b(·) are continuous functions defined from I = [α, β] to R
and t0 ∈ I .

Integrating factor. Find a differentiable function µ : I → R such that the


ODE can be written in the form
d
(µ(t)y (t)) = µ(t)b(t), y (t0 ) = y0 .
dt
Then integrating in both side, we can get the solution.
Step 1. Multiply the ODE with µ and obtain

µ(t)y 0 (t) + µ(t)a(t)y (t) = µ(t)b(t).

Step 2. Find µ(·) such that it is a positive, differentiable function satisfying

µ0 (t) = a(t)µ(t), ∀ t ∈ I , µ(t0 ) = 1.

Rt
Set A(t) = t0 a(s) ds for all t ∈ I . Since a(·) ∈ C (I ), then A ∈ C 1 (I )
(Check) and A0 (t) = a(t) with A(t0 ) = 0. Thus, taking µ(t) = e A(t)

d A(t)
(e y (t)) = e A(t) b(t),
dt
and integrating over [t0 , t] in both side
 Z t 
y (t) = e −A(t) y0 + e A(s) b(s) ds , ∀t ∈ I.
t0

Note that y (·) is defined for all t ∈ I and is C 1 (I ) and y (·) satisfies the
linear ODE.
Examples and properties

Ex. Consider a(t) = a positive constant for t ∈ R and b(·) ∈ C (R) such
that |b(t)| ≤ Me −at for all t > 0. Solve

y 0 (t) + ay (t) = b(t), y (0) = 1.

1. Does the solution exist for all t ∈ R?


2. Is the solution unique?
3. Describe the asymptotic behavior of the solution as t → ∞.

Ex. Solve: ty 0 (t) + y (t) = 0, y (1) = 2. Does the solution exist for all
t ∈ R?
Ans. No, the solution cannot be extended beyond t ≤ 0. a(t) = 1t has a
singularity at t = 0.

You might also like