0% found this document useful (0 votes)
12 views30 pages

MA 108 - Ordinary Differential Equations: Suresh Kumar

Uploaded by

icarus
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
0% found this document useful (0 votes)
12 views30 pages

MA 108 - Ordinary Differential Equations: Suresh Kumar

Uploaded by

icarus
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/ 30

MA 108 - Ordinary Differential Equations

Suresh Kumar

Department of Mathematics,
Indian Institute of Technology Bombay,
Powai, Mumbai 76
[email protected]

April 4, 2024

1/38
Outline of the lecture

Method of variation of parameters


Method of undetermined coefficients
Non homogeneous Cauchy Euler DE

2/38
Variation of Parameters
We consider the non-homogeneous linear ODE
Ly := y (n) + p1 (x)y (n−1) + . . . + pn (x)y = r (x).
We proceed exactly the same way as n = 2.
Let the general solution of the associated homogeneous DE be
y = c1 y1 + c2 y2 + . . . + cn yn .
Method of variation of parameters suggest particular solution of
the form
yp = v1 y1 + v2 y2 + . . . + vn yn (∗)
We use the following ansatz ((n − 1)th order contact conditions).
v10 y1 + . . . + vn0 yn = 0,
v10 y10 + . . . + vn0 yn0 = 0,
··· ·
(n−2) (n−2)
v10 y1 + ... + vn0 yn = 0.
3/38
Variation of Parameters
Differentiate (*), we get

yp0 = v1 y10 + . . . + vn yn0 + v10 y1 + . . . + vn0 yn

Using the Ansatz, we get

yp0 = v1 y10 + . . . + vn yn0 .

Differentiate again and use the anstaz, we get

yp00 = v1 y100 + . . . + vn yn00 .

repeat this, we get

......
(n−2) (n−2) (n−2)
yp = v1 y1 + . . . + vn yn .
(n−1) (n−1) (n−1)
yp = v1 y1 + . . . + vn yn .
4/38
Variation of Parameters

Differentiate the above, we get


(n) (n) (n) (n−1) (n−1)
yp = v1 y1 + . . . + vn yn + v10 y1 + . . . + vn0 yn .
(n)
Now substituting the above equations for yp , yp0 , · · · , yp below,
we get
(n) (n−1)
r (x) = yp + p1 (x)yp + . . . + pn (x)yp
= v1 L(y1 ) + v2 L(y2 ) + . . . + vn L(yn )
(n−1) (n−1)
+v10 y1 + . . . + vn0 yn
(n−1) (n−1)
= v10 y1 + . . . + vn0 yn .

5/38
Variation of Parameters
Thus we have,
 0
y1 y2 · yn
   
v1 0
 y10 y20 · 0
yn   v20
    0 
=
  · .
 
 · · · ·  ·
(n−1)
y1
(n−1)
y2
(n−1)
· yn vn0 r (x)
Use Cramer’s rule to solve for

v10 , v20 , . . . , vn0 ,

and integrate to get the formulae for

v1 , v2 , . . . , vn ,

and a particular solution

yp = v1 y1 + v2 y2 + . . . + vn yn .

6/38
Geometric interpretation of the Ansatz
As in the second order case, let Y (x, ξ) denote the solution of the
modified non homogeneous equation by ’killing’ r (x) from ξ
onwards.
Then Ansatz is given by the (n − 1)th order contact conditions of
the curves yp and Y (·, ξ) at the point (ξ, yp (ξ)). The (n − 1)th
order contact conditions are: at x = ξ

Y (x, ξ) = yp (x)
Y 0 (x, ξ) = yp0 (x)
··· = ···
(n−1) (n−1)
Y (x, ξ) = yp (x).

Now by substituting

Y (x, ξ) = v1 (ξ)y1 (x) + . . . + vn (ξ)yn (x)

into the contact conditions and proceed as in the case of n = 2 we


get the ansatz.
7/38
Example

Solve
y 000 − y 00 − y 0 + y = r (x).
Characteristic polynomial for the homogeneous equation is

m3 − m2 − m + 1 = (m − 1)2 (m + 1).

Hence, a basis of solutions is

{e x , xe x , e −x }.

We need to calculate W (x). Use Abel’s formula: (here


p1 (x) = −1)
Rx
W (x) = W (0)e − 0 p1 (t)dt
= W (0) · e x .

8/38
Now,
ex xe x e −x
W (x) = ex e + xe x
x
−e −x .
ex 2e x + xe x e −x
Thus,
1 0 1
W (0) = 1 1 −1 = 4.
1 2 1
Hence,
W (x) = 4e x .

9/38
Variation of Parameters

Set
0 xe x e −x
W1 (x) = 0 e x + xe x −e −x = −r (x)(2x + 1).
r (x) 2e x + xe x e −x

Obtained by replacing first column of W (x) by (0, 0, r (x)T .


Similarly,
W2 (x) = 2r (x), W3 (x) = r (x)e 2x .
Note that
  0   
y1 y2 y3 v1 0
 y10 y20 y30   v20  =  0  .
y100 y200 y300 v30 r (x)

10/38
Variation of Parameters

Hence
Wi (x)
vi0 = , i = 1, 2, 3.
W (x)
Therefore,
x x x
−r (t)(2t + 1) r (t)e 2t
Z Z Z
2r (t)
y (x) = e x
dt+xe x dt+e −x dt.
0 4e t 0 4e t 0 4e t

11/38
Annihilator Operator-revisited
If A is a linear differential operator with constant coefficients and
f (x) is a sufficiently smooth differentiable function such that
A(f (x)) = 0,
then A is said to be the annihilator of the function f (x).
Examples :
1 D 4 annihilates 1 − 5x 2 + 8x 3 . (i.e., it is a solution of DE D 4 y = 0.)
2 D + 3 annihilates e −3x .
3 (D − 2)2 annihilates 4e 2x − 10xe 2x .
4 D 2 + 16 annihilates cos 4x, sin 4x or any of their linear
combinations.
5 D 2 + 2D + 5 annihilates 5e −x cos 2x − 9e −x sin 2x.
6 D n annihilates 1, x, x 2 , · · · , x n−1 .
7 (D − α)n annihilates e αx , xe αx , · · · , x n−1 e αx .
8 (D 2 − 2αD + (α2 + β 2 ))n annihilates e αx cos βx, xe αx cos βx, · · ·
x n−1 e αx cos βx, e αx sin βx, xe αx sin βx, · · · x n−1 e αx sin βx.
12/38
Example 1

Using Annihilation argument, find the form of the particular


solution and then compute a particular solution of

y (4) − 16y = x 4 + x + 1.

Here,
L = D 4 − 16 = (D 2 + 4)(D − 2)(D + 2),
and
A = D 5.
Basis for Ly = 0 is {e 2x , e −2x , cos 2x, sin 2x}.

AL = D 5 (D 4 − 16).

Hence basis for ALy = 0 is


{1, x, x 2 , x 3 , x 4 ,e 2x , e −2x , cos 2x, sin 2x}.

13/38
Annihilator Method

Any general solution y = y1 + yp , where y1 is the complementary


function and yp a particular solution is a solution of ALy = 0.
Hence we look for a solution of the form

yp = A1 + A2 x + A3 x 2 + A4 x 3 + A5 x 4 .

(D 4 − 16)yp = x 4 + x + 1 =⇒
24A5 − 16(A1 + A2 x + A3 x 2 + A4 x 3 + A5 x 4 ) = x 4 + x + 1.
Comparing coefficients, get:
1 1 5
A5 = − , A4 = 0, A3 = 0, A2 = − , A1 = − .
16 16 32

14/38
Example
Solve the DE:
Ly = (D 2 − 5D + 6)(y ) = xe x .
Characteristic polynomial for the associated homogeneous
equation is
m2 − 5m + 6 = (m − 2)(m − 3).
Basis of Ly = 0 is {e 2x , e 3x }. Take,
A = (D − 1)2 =⇒ AL = (D − 1)2 (D − 2)(D − 3).
Basis of ALy = 0 is {e 2x , e 3x , e x , xe x }.
So yp = Ae x + Bxe x , i.e., we need to find A, B such that
L(Ae x + Bxe x ) = xe x .
This gives (2A − 3B)e x + 2Bxe x = xe x . Thus,
3 1
y = e x + xe x
4 2
is a particular solution.
15/38
Example

Find the candidate solution of

y (4) + 2y 00 + y = 3 sin x − 5 cos x.

The auxiliary equation is m4 + 2m2 + 1 = (m2 + 1)2 = 0, and


therefore a basis of solutions to Ly = 0 is

{cos x, sin x, x cos x, x sin x}.

Let A = (D 2 + 1). Basis of solution of


(AL)(y ) = (D + 1)(D + 1)(D 2 + 1)(y ) = 0 is
2 2

{cos x, sin x, x cos x, x sin x, x 2 cos x, x 2 sin x}

So a candidate of particular solution of the non-homogeneous DE


is y = x 2 (A cos x + B sin x).

16/38
Annihilator Method

Exercise: Get candidate solutions by the annihilator method:


1 y (4) − y (3) − y 00 + y 0 = x 2 + 4 + x sin x.
2 y (4) − 2y 00 + y = x 2 e x + e 2x .

17/38
Undetermined Coefficients
We have seen the method of undetermined coefficients for the
second order case. For the higer order case also, form of the
particular solution is as in the second order case.
r (x) = yp =
n ax
x e , a ∈ R and a is not A0 e + A1 xe + · · · + An x n e ax
ax ax

a root of the charact. eq.


x n e ax , a ∈ R and a is x µ (A0 e ax + A1 xe ax + · · · + An x n e ax )
a root of the charact. eq.
with multiplicity µ
X n X n 
x n e ax cos bx / x n e ax sin bx e ax Ak x k cos bx + Bk x k sin bx
k=0 k=0
a + ıb is not a root
the charact. eq.
n
X n
X 
x n e ax cos bx / x n e ax sin bx x µ e ax Ak x k cos bx + Bk x k sin bx
k=0 k=0
a + ıb is a root of
charact. eq. 18/38
Example 1
Find a particular solution of

y (4) + 2y 00 + y = 3 sin x − 5 cos x.


1
h(x) = − (5 + ı3)e ıx ⇒ h(x) + h(x) = 3 sin x − 5 cos x. Set
2
Ly = y (4) + 2y 00 + y . The characteristic equation is
m4 + 2m2 + 1 = (m2 + 1)2 = 0, and therefore a basis of Ly = 0 is
{e ıx , xe ıx , e −ıx , xe −ıx }.
D − i annihilates h(x). Basis for (D − i)Ly = 0 is

{e ıx , xe ıx , x 2 e ıx , e −ıx , xe −ıx }.

Form of the particular solution for Ly = h is Y = Ax 2 e ıx .


Substituting, we get:
1 1
8A = (5 + ı3) =⇒ Y (x) = (5 + ı3)x 2 e ıx . Therefore
2 16
5 3
yp (x) = 2Re(Y (x) = x 2 ( cos x − sin x).
16 16
19/38
Example 2
Find the candidate solution for

y (4) − y (3) − y 00 + y 0 = x 2 + 4 + x sin x.

Note that the characteristic equation is

m4 − m3 − m2 + m = m(m − 1)2 (m + 1) = 0.

Thus a basis of Ly = 0 is {1, e x , xe x , e −x }.


For x 2 + 4, form of the particular solution is x(A0 + A1 x + A2 x 2 ),
since m = 0 is a root of the characteristic equation with
multiplicity 1.
For x sin x, the form is

(B0 + B1 x) cos x + (C0 + C1 x) sin x.

So our final candidate is

x(A0 + A1 x + A2 x 2 ) + (B0 + B1 x) cos x + (C0 + C1 x) sin x.


20/38
Example 3
Find the candidate solution for

y (4) − 2y 00 + y = x 2 e x + e 2x .

Note that the characteristic equation is

m4 − 2m2 + 1 = (m − 1)2 (m + 1)2 = 0.

So a basis of Ly = 0 is {e x , xe x , e −x , xe −x }.
e 2x on the right hand side suggests

Ae 2x

and
x 2 e x suggests
x 2 (Bx 2 + Cx + D)e x ,
since m = 1 is root with multiplicity 2. So final candidate would be

Ae 2x + x 2 (Bx 2 + Cx + D)e x .
21/38
Review Question 1

For the non-homogeneous equation y 00 − 5y 0 + 6y = x cos x, the


form of yp using the method of undetermined coefficients is
————-.
(a) A1 x cos x + B1 x sin x
(b) (A0 + A1 x) cos x + (B0 + B1 x) sin x
(c) x((A0 + A1 x) cos x + (B0 + B1 x) sin x)
(d) None of the above.

Solution : 4 (b)

22/38
Review Question 2

1 0 2
Solve the IVP : y 00 + (1 + )(y ) = 0, y (0) = 1, y 0 (0) = 1/e.
y
dv dy dv dv
Put y 0 = v . So y 00 = = =v
dx dx dy dy
dv 1
The DE becomes v + (1 + )v 2 = 0.
dy y
Solve for v : v (y ) = Ce −y y −1 .
Use initial conditions to obtain C = 1.
Thus ye y v = 1. Solve to obtain ye y − e y = x + C1 .
Use y (0) = 1 to obtain C1 = 0.
Hence, (y − 1)e y = x .

23/38
Review Questions 3 & 4

Solve using the method of variation of parameters :

(x 2 + 1)(y 00 − 2y + 1) = e x , y (0) = y 0 (0) = 1.

24/38
Cauchy Euler equation-Calculus with xD

(xD)k := (xD)k−1 (xD), k ≥ 2.


(xD)2 y = (xD)(xDy ) = xDy + x 2 D 2 y ⇒ x 2 D 2 =
(xD)2 − xD = xD(xD − 1).
x 3 D 3 = xD(xD − 1)(xD − 2), etc
Observe that (xD)k (x m ) = mk x m , k = 1, 2, · · · . Hence for
any polynomial P(x), P(xD)(x m ) = P(m)x m . This leads to
the indicial equation of the Cauchy Euler DE.
If α, β are the roots of the auxilliary equation
m(m − 1) + am + b = 0, then

(xD−α)(xD−β)y = x 2 D 2 y +axDy +by = (xD−β)(xD−α)y .

25/38
Non homogeneous Cauchy Euler equation

Solve x 2 y 00 + 2xy 0 − 6y = 10x 2 .


Set
L = x 2 D 2 + 2xD − 6.
The indicial equation is

m(m − 1) + 2m − 6 = m2 + m − 6 = (m + 3)(m − 2) = 0.

1 2
Basis for Ly = 0 is { , x }.
x3
Need to find out an appropriate annihilator. Notice that D 3 is an
annihilator for r (x) = 10x 2 . But we don’t know how to find a basis
for D 3 (x 2 D 2 + 2xD − 6)y = 0, since it is neither with contstant
coefficients nor a Cauchy Euler ODE.
Note that A = (xD − 2) annihilates 10x 2 .

26/38
Cauchy Euler equation

ALy = (xD − 2)(x 2 D 2 + 2xD − 6)y = 0


is Cauchy Euler.

AL = (xD − 2)(xD(xD − 1) + 2xD − 6)


= (xD − 2)((xD)2 + xD − 6)
= (xD − 2)(xD − 2)(xD + 3).

1
So the basis for ALy = 0 is {x 2 , x 2 ln x, 3 }. Thus, yp = Ax 2 ln x.
x
Substituting yp into the ODE we get

(xD − 2)(xD + 3)Ax 2 ln x = 10x 2 .

Therefore 5Ax 2 = 10x 2 , i. e., A = 2. Hence yp = 2x 2 ln x. The


general solution is y = c1 x −3 + c2 x 2 + 2x 2 ln x.
27/38
Cauchy Euler equation

If (x − m)2 denote the auxilliary equation of


x 2 y 00 + axy 0 + by = 0, x > 0, then the basis for its solution is
{x m , x m ln x}. Hence
(xD − m)(x m ) = 0, (xD − m)2 (x m ln x)) = 0.

28/38
Cauchy Euler equation-Example 2

Solve Ly = x 2 y 00 + 4xy 0 + 2y = 2 ln x, x > 0

Ly = (x 2 D 2 + 4xD + 2)y
= (xD(xD − 1) + 4xD + 2)y = ((xD)2 + 3(xD) + 2)y
= (xD + 1)(xD + 2)y .

1 1
Basis of Ly = 0 is { , 2 }.
x x
(xD)2 annihilates ln x.
1 1
Basis for (xD)2 Ly = 0 is { , 2 ,1, ln x}
x x
Form of the particular solution is yp = A + B ln x.
3
Substitute yp is the DE, we get A = − , B = 1
2
3
Therefore a particular solution is yp (x) = − + ln x.
2
29/38
Cauchy Euler equation-Example 3

Solve Ly = x 3 y 000 + 4x 2 y 00 + xy 0 − y = x, x > 0


Auxilliary equation is m3 + m2 − m − 1 = (m − 1)(m + 1)2 = 0 ⇒
1 ln x
Basis for Ly = 0 is {x, , }.
x x
(xD − 1) annihilates x. So
1 ln x
Basis for (xD − 1)Ly = 0 is { , , x, x ln x}.
x x
Form of the particular solution is yp (x) = Ax ln x.

A 000 A
yp0 = A + A ln x, yp00 =
, yp = − 2 .
x x
1 1
Substitute in the DE, we get A = . yp (x) = ln x, x > 0.
4 4

30/38

You might also like