MATHEMATICS III -

(Differential Equations)

Kaushik
Dwivedi
 MATHEMATICS -

III

Chapter -
1 :
Preliminaries of
Differential equations
Chapter -
2 :
first order Differential equations
Chapter -
3 : second order Differential equations
Chapter - 4 :
qualitative behavior of solution

Chapter -
5 : power series solutions and special
functions .

Chapter -
6 : Fourier series

chaplet -
7 :
Boundary value
problems
Chapter -
8 : some
special functions
chapter -
9 :
Laplace Transforms
and inane
Chapter -
to : One dimensional heat

equations
 Preliminaries of Differential Equations

Differential equations - :

variable and its

They involve dependent
one

derivatives with respect to one or more independent

variables .

highestordadaiuatiue)
Order and
degree
- :

↳order of the

The power
or enponent
order derivative
of the highestthe
in DE
occurring
.

Linear DE -
:

↳ When the dependent variables and their derivatives

occur
first degree and arent multiplied together
.

in
 First Order Differential Equations
A first order DE is of the form -
:

gln.y.gl/-- 0

↳ '
fcn g) canonical form
y = ,
→

order DE
first
- :
Methods
for solving any
1. Variable
separable DE - :

separable form
as
It is written in variable
'

y
-_
Fcs Gly )
OR

dy_ = Few dn
only )
integration
On .
we have

faddy ,
-

fans dnt c

DE
Homogeneous
:
2. -

be homogeneous
is said to
flu
A '
DE g)
y
-
,

if fltn ty ) flu y) To solve it we use

,
=
,
.
,

VK
y
-
3. DE of the
form y Can + by -1 c) / (pn&qyer )
'
=

we transform this into

×"
%, In
;÷÷ )
- a -

Yak
y __

Enact DE - :

The first order DE be

deg flu g)
: = can
,

written in the canonical form :

Mln g) dnt N(
my)dy=0
,

where
flmy)
M£Yn→y
=
-

,
It if Mdn - Ndy
enact is
is said to be
an

enact differential of some function

f- Cny) that ,
:

df
Mdm &
Ndy -_

Enact DE
Necessary andsufficient condition for - :

MCR, g) dn -1 Ncr g) dy ,
= 0 is enact iff :

8dg :-.
=
Proof :

Assume that
mln g) du + Nln g) dy
, , = 0 is exact
Then,
by definition we have ,
:

Mdr +
Ndy =
F

doin du +
81g dy -
df

: have
Yy
we
M Of
.

- and N =

0
2m
8

Both def
g- 8¥oi!÷=8÷y
§n¥gy
and continuous
are ,
therefore they
are
equal .

Hence ,

Iffy -

%n ^

on
integrating ,
we
get
:

F- fmdn -1g (g)

%y=§-y fmdn -
gigs
 N
Egg fmdn
'
&
g ly )
=

gigs
=/µ %gfmdn)dy
-

I 1
do

function of y only
.

Is a

% (N -

8-yfmdn) - 0

doin §÷ yfmdn
-

,
= °

dong dingy
or - -_ 0 -

This completes the proof .

Solution Enact DE
of
- :

Mdn
If +
Ndy = 0 is enact then , E- Ln , g) = c
heads as

fmdn &
kN -8g fmdnn) dy - e
Existence and Uniqueness of Integrating factor :
-

Enis fence is ensured provided a solution for

Mdn -1
Ndy -_ 0 enislt .

Integrating factor µ = ethlntdn

In ( omfg Ign )
where has - -

order linear DE
Integrating factor of first
- :

y 't pray =qw

has solution :

[email protected]
eduction
of order - :

by reducing
be solved the
A second order DE can

order order DE
or to
first .

Substitute y'-_p-
 Second Order Differential Equations
Second order LDE - :

LDE of the form

Any second order
:
is
-

flay y y ) ,
'
,
"
= 0

written in
general form as - :

" '
y + party + qcnsy =
red

linearity pewneiple :

If y ,
and
yz
are
any
two solutions of the homogeneous
IDE " ' then aye cage is also
y + prey +
girly = 0
,

Sol 9 and ez
for any
a .
.

Linearly independent and

linearly dependent functions -
:

not constant
f w and gems are LI if fly is a

multiple of gold .

Wronskian :

W ( Yi yz) Y' 82
, = -
Gita yagi
-

yi yi
Remark :

Two solutions LDE ate

yz of homogeneous
and
ye a

linearly dependent iff w(y y a) 0 A , , = re c- [ b)

a.

Result - :

The Wronskian of two solutions y , @ and yin

defined on [a b) of a homogeneous LDE
,

identically zero
"
or never
Y t either
'
p y →
qcn , is

Zero

proof :

"
since and solutions of plus y
'
y ,
y~
are
y
t +
qcnsy -_ 0
,

hme.gg
we

*
pen, y
'
→ gens y - O → ①

gilt pins y:& guy ,

= o - ②

Multiplying ② by ye and ①
by y,
and subtracting ,

Yi yi plus Lyeyi yugi)

" 0
yay *
-
- -

d¥n + pens w - o

w -
ce
-

fPHdn
W never vanishes
W is 0 if l is 0 otherwise
.

i.
Use
of known solution to find another :

If solution is known , then the other solution

a
y
yz can be found at
ya Vey
=

e- 1PM
V =

flagg dndn
IRN du
ye f Lyy
°"
e- du
Yz =

and then have

we another
ay →
czy ,
as
solution
general .

with constant coefficients

Homogeneous LDE -.
:

" 0 → I
y + pyo → qy-_
We have
let em be solution ,

y
= a
.

(m2 1- pm g) em
"
+ __
0

i. m ⇐ p ITH
-

solution
Then we
get
the
general
y
=
Gemma ↳ email
If my = a + ib and
mz = a -
ib
,

elatibl = can ( cosabut isinbn) .

The solution
of ①
is
general
e- P%
g- (Gd Can)
Transformation to
homogeneous LDE with constant
coefficients :

"
y -
pray
'

qing e __ 0

Using the
transformation E-
fivqndn ol
z1=%
"2
If ⑨ & 2pq) / q is a constant .

Remark :

The "
equation sty +
pay
'
+ o where p q are
qy
-
,

are constants reduces to

dty + (p 1)
ddytzoqg -0
-

Ii
.

which is
again a
homogeneous LDE with constant

coefficient .
General solution
"
of a
nonhomogeneous IDE - :

'
y -
pay tqwy-r.tn → 1

The
homogeneous LDE
y
"
t
ping '&q☒y-_ 0 has
a solution
yµ =
Uy, + Czyz .

Also I must solution

have a
particular yp .

Then , The overall solution of z is

y -
yn - yp
Method of undetermined coefficient :

rent and use

Tn 1
we analyse the
function
particular
,

Kail method
a

hit and
to
form
solution
yp .

Method of variation
of parameters :

solution
This method is used to
find a
particular
DE 1
Yp of the non
homogeneous .

be the solution of the

let 4yd ↳ y,
y
=

corresponding homogeneous equation

Then we substitute Visit Uzyz
y
,
-_
Where v @
,
and v24) follow viyi a vagi - red
and ,
Vi yikes
= -

& vi. =
yir
¥421 wlyi , Ya )
Therefore this leads to

y Ye
f-yzkm-dn-yzfyerd-dnwly.ge
-_

) Wlyr yi
,

Operator methods :

we denote the differential operator ddg by D.

that and
dy_dn-_ Dy
'
such
y -
,

daffy
" "
y __ =
By .

Therefore
the DE
y t pytqy
- ray
can be written as

CDH pD&q)y - r @

[email protected]
Oh,

rent , which
1µg
We have is a

y
-_
.

solution of the DE
particular .
1) If a is a constant and f-(D) = D- a
,
then :

y ( D- a)
Iga
- am a @
y -

dy_dn -

ay / - rn , or
y -_ can n@ e-andre
,

[email protected]
thus ,
1. has -

-
a

Remarks : The use

of partial fractions comes in
handy
methods
while
solving using operator
.

write
2) If he @ is some
polynomial in then r, we

powersin ascending
the series
enpansion of 1-
of D .
f. (D)
3) If K is a constant and ray -
ekkgcns ,
then

Let"g☒)
g-
empty god
- .

4) constant, and hey et "

then
If K is a =
,

Y-tf-jtk-gp.ge
"
? In case
fat -0 ,

we have

"
¥,
em
g-
refile
-
 Power series solutions and Special Functions
--

Basics of power series - :

An
infinite series of the
form :

É am
( n -
no )^__ ao day CR -

Rotas (n -
ko Tt . . . .

N I 0

If f-(n) - § an ( n -
ko )n__ ao t ar (n -
not
n- 0 "

Az ( N - Ro ) t . . . .

as

fl K) -
[ man ( n -
ko
) 1-
"

ay & zoy ( n -
not t . . . .

n = 1

and so on . . .
.

Power series solution -

:

and
Ordinary single points :
-

"

y t
pay
'

qcnsy -0 &

If the
functions pay and qing analyte at
are n -
ro
,

then n =
ko is called an
ordinary point of the DE .
If pH at but
and
qca) fail to be analytic re

CR at then
-
no )
pens and @ -

scope qcn, are analytic Ro

,

regular singular point

do called
-

is a

Frobenius power
series solution -
:

Theorem :

"
& pens
Y has
regular point
'
+
qcnsy
o → as
y no a
- ,

then there crusts at least series solution

one power
as :
her
§ an Cn -
)
no

y =

n= 0

where he is the rod of the quadratic equation

rit Lpo Dnt go
- -_ 0 .

and , him @ ) plus

po no
-
=

↳ no
and

go ⇐
him G- not glad
u → no
Gauss
Hypergeometric
'

Equation - :

Ncn 1) y effort b- 1)
"
-
re -

☐ y' +
aby - o

↳
hypergeometric equator
.

This has atleast solution of the

one
frobenius
form :

E Ñ(
" "

aotqrtazrbas.se?n--
an n

y
- =

0
+ . . - .

)
 SOME SPECIAL FUNCTIONS

1 .

Legendre polynomials
Rodrigue formula
's
-

Orthogonality
-

Legendre series
-

2 . Gamma function
3 .
Bessel functions
-
Second solution of Bessel 's DE
-

Properties of Bessel functions

-

Ortho
gal properties of Bessel functions
Fourier Bessel
-

series
Legendre Polynomials -
:

HE 1) "
2mg nlntily
'
y t 0
-
-
=

Finding PS solution for n - 1

CR 1) Ln -11) y "
-
t
2kg1 - mln -
Dy - o

put
t-L-A-a-II-tg.tn -
D

i. tlt 1) ÑY_ lot 1)

dyµE
-

a
nastily 0
-

=
"
dt
-

a 4641 =3 c- 1, at =
-
ncntl)
a- -
n
, b = In -113

YI F f n n +1 1
b)
-1-2 (
-

, n -

, ,

Elaine ,m=
;÷- aY¥Y¥¥_ñ
1- a
-

& - - - -
- - .

nlY¥ (1--2)
i.
y l t -

Pncnt
-
+ . . - .
=

Now , when n
-40, Po Cn) - 1

PI CH l this, all terms

-1,71-11-2-1 after
-

,
are 0 .
 i. Pn Cn) is called Legendre 's polynomial .

Pn ( n) It
ngln.gl#(i-zn-)+-ntntDlntDln-4
-

-11%-7
f. - - --
l 2 1
. . .
2

Note : Pn (1) =L

's
Rodrigue

ddn-nnflnt-HYPok-L-o.co
formula Pride
-

In ,

BCN
Iggy
=L : - - 2k

,
and so on

Proof of Rodrigue's formula - :

let
@ Dn
v-_ 2-

I
2mn ( x2
n
1)
-

V1 = -

i.
@2-1) v
,
- 2mn v - 0

Diff City times

( VV) n
- Uvn -1 hey V1 Unit . . . .
→ Unr
 (22-1) v1 - 2mn V - O

= :
CN D

vnez-tln-DGHVnm-tlnl-Dn-n-i-2nf.se
-

Vn ,, + n-
( Dun
] -_ 0

(R2 1)
-

Une z & 2N Unh -

nlntl) Un - O

Un is a solution of the

legendre DE -

i -

Un is a
polynomial of degree n .
v-_ -

]
µ yn

Pm (n) -
Cvn =
Cdn Latin
Iden
Pnh
cdgn-nnf.cn
-1
1) In -D ]
- "
-

R- L
- n ! ↳ + 1) "

(tuns
D)
e +
carrying a -

N-4
I →
C in ! .
zn

c- 1-
2h .
n !
 Po @ =L Py @ Point
Ig ( 3ns
-n
, ,
__
-
1)
Pz @ -
Iz ( 5h3 -
3m)

Note :
Any polynomial can be written
using
Legendre polynomials
's .

Pnm
¥ ÷nn(ñ
^
=
-

1)
.

Also , Pnls) =L

Orthogonality
e)
-
Im @
I
Pn @ dn = 0

ii pile dn = 2
/ (2n -11
-

Proof :
M
i)
Pmlny ti-

Dy
"
→
2mg
'
mCm&Dy o
y
-
-
-

z = Pn (a) =
42 1) z" + znzl
-
-
ncn-1117--0

↳ 2
 ② .

2- -
②y HE 1) ly"z "
g) 42k1g't -21g)
-
-

z -

+ @ ( nm) - m
1m41)]y 2- = 0 .

Ida ④ g)IN D) + (
z -
'
2- -
n -
m
) ( name 1)
yz
integrating from 0 to "
1:

( name 1
! / gzdn
0 +
(n -
m ) = 0
,
'

-
I
/ pm (a) Pncrjdn = 0

ii) *Lecture Nolet *

Legendre series :

[email protected]
f@
=
. .

1- Cn Pn @ + . . -

"

/ Pn @ fcnsdn
"

th / pick dk
tansy
Au other
-1 ,
cancel out
↳
2%+1
=

"

/ Pnd ftndn
an
2h22
-_
)
,

Eg d- sent E) I §
-
"
+
tnpncn
:

te n - o

Generating function
of
Legendre polynomials

[email protected]
- in Pnb) =
lntypneitnpn.int
eg :
find the value of Pznlo)
solution I
É
:
"

= t Pncn,
✓ 1 -2kt -1+2 n= 0

At
-
"2

-_g÷→=
R 0, LHS
__
d- E)

= 1-
1- the 1-23-2 tf
RHS =
Polo) ÷
+
pilot + Blot 't - - - -

i. Paulo) =
"
C- 1) 1.3 . . . -
12h13
" 2h n !
eg :-, / Ps @( n
-

'
-
2m41)dn=
?

I , / Isin )( a BIN -10 Palate Phased Pork)dr

If ! :| !
'

I / Psh Pztrdn
=
- - → . .
. .

All
integrals
are
equal too .

as
,
/ Ñmlnpnsydn - o
Gamma function - :

cry -

of 2- Kum da ,
n > 0

Tcm) =
@ 1) In D
T(1) =L
-
-

5151--4 ! :
5€) TT -

G
of %
:

THE) a- 112dm
/ %-8dg
"
=
2 "
Ey
=
in
,

El:) ] : :|:[e-aimdndy
Put

r s i n o d n d y - rd rd o - H Y. i e
reoso ,
a -_
y =

HEID
- "

"

- t

=
FT
 when u
n
so :
Io
fln) -_
of e-
✗
sent dn

when n 20
,
and not an integer ,

Mn) = Tht D
a-
If n = 0
,
-1
,
-2
,
. - . -

In) = A

eg : T

T&&,-_
=
-2 't

eg Tff) 9-1-11-312=-5 't

:
-

Bessel Functions - :

"
Ny try pYy
'
- Cri - - o

regular singular point

-

n- o →

y § Anant ? r =P
=
,
-
P ,
an =
-an-z_
0
n
@ →e) 2- p2
=

For r =P

aoxPÉ
,

(-13^412)
"

y -

n - o n ! ( Pti) ( ped . - . . -

Lptn)
Put
Ipp
ao =
'
!

2h - P
Tp
"
=
I C- D ( a/ 2)
n ! T ( nap+1)
I Bessel 's
function of first kind

Now '

YI Y÷g
To (a) = 1- - -
- - - - - -

I@
Iz ,¥z , @P
=
-
- - - - - - -

Then we have :

gotµ, = -
J, W

Normal form

C- -1£ )
"
a + I
u -_ 0

equation
:
The solution of the Bessel 's

y = 4 Jptr & G 5pm - 1

Now '
T.pcxj-L-s.JP Jp@ ,
when p is an
integer .

i. I cannot be a solution when p is an

integer
 second solution :
Constructing a

Yp Cn) =
Jp @ lost -

Tpm → finite
simpT p → mtz

when
p is an
integer :

y = a 3pm → ↳
yptkl
de
of the
Bessel function
second kind

Properties of the Bessel

function :
-

[ xp Tph] XP Jp , (a)
1)
%
=

" "P
Proof : "

(E)
KP Jpcn) = up § C- e)

1Tntp-DRPJp@y_zl-iYn2nt2pzMtPnl.T
n
n
-
- o

( n -p - 1)
Now ,

d_dnCRPJp@jf-Ec_pnC2n-zpjn2n-2p-1zM-Pn1.T In &P&1)
 [email protected] Ln&P)
:

dqn[ xp Tpm]
.

RP
=
Jp-104
2)
dd_n[ x-P Jp (a)] __ - x-P Jpn ( n)

3) Tp ' @ -
t.f.Jp.im Jp→W]
-
→ from 142

A) Jpg, (a) JPY4 Tp in → recurrence

2dg
-
-

= relation

Corollary - :

1)
/ RP Jp -
,
(a) dn - xPJpW&C

2)
fze-PJpmcrgdn-s-k-PT.pk )-e
Note :

511241 =

⑨q Sink
( two series are equal )

2)
J¥n
J-

11241 = lose

3)
53124--1-25112 -5-112 -

☒n(8in÷ - cosh
)
from Jp,, Tp Tp
=
2Pa i
-
-

A) Ip CD =
C- 1) PJPCRJ

Orthogonal properly :
-

If Am and Xn are two zeros of Jp then

,
:

/ Tl Jpltmk Jp Ann) die

{ & Jp:
Q m *n
⇐
o
( tn) man
,

Fourier Bessel series - :

Any function flu) can be written using

the

Bessel
functions .

flat - É an Jplilnoe)
n- I
 ofTeJpHnnJf
of =
@ dn-I z JpI"

n
i n Jan. i
Jp Hn
. a
n
n ) fcnsdn

d- Tpf ,

as

eg :
I

n-zzy-n-g. a m?0&nH1--
=

÷ ,
our
Joann

an -

¥µf%;÷ Toldnntdn

2-
94in [¥27 niff
¥5 Lin
=

,
 LAPLACE TRANSFORM
f@ ,
n c-
[a. b)
[email protected]) finds -
FLP)
a- O, b-
e- Pm
w
, f. =
Lp
,
nj =

Laplace transform - :

L [ few] -
of E- Pnfensdn =
Flp)

Inverse Laplace -
:

'
1- [ Flp)] -
f@
transform
Linearity property of Laplace - :

I
Lafcns & legend -
attend to LEAD
 Properties of Laplace Transform - :

1) I [ can] I 2) I [ rn]
= =
Until
p
¥
-
a
.

3)
LEinanJ-pa-azafhfosanT-pp-o.rs
5) Lffinh an] a- 6)

l@shanI-pp2_a2Gj.L
=

phase

[ siren]
L[ Sirin] [ _¥n]= :-( LET
=L '
-

Leonid
1. ( & ÷*)
-
-

G
m[¥i=e
'

[¥i-_¥
: "
:[ "" "
;
G
G i
#pig [ Fire ;] [¥
:
i
¥]
-
- " -

= R -
Sinn
Enis three of Laplace Transform
- :

it fry is
piecewise constant
)
ii
f f☒1 I Mean →
exponential order

then Lffey] enists

for p > c ( if em] =
Ffp,

If L
[fend = F
, then

1) I @ fed
am
= F- (p -
al

2)

Lf ikD-pFlp1-f@Lf lCnJf-p2Flpg_pfl0J-fl 0J31Lf.of

"

fltldt] -

F%
Proof :

"

let
glory -
of fltdt
-7
LEG @D= tiff XD
'
 p l§Ñ -

g @) =
FCP)
LIGHT
FYI
=

g) I[
Rf@D= C- 1)
dip
Proof :

of e-
Png @ dn =
FCP)
sides wit p
Differentiate both

of In e-M fensdn = Flp)

of IM [ nfcnjdn - C-2)
dip
-

1Er find =
C- 1)
dFqp
5) L[
nappy] -1-17%4
6)
t.CH#J--pfF-ltdt
proof :

let
fl -
girl
 Lff xD - L
@ gas] -
off xD
1- 1) G ' Lp)
=

=
Fcp) GYP) = -

Flp)

p
/ Tip)dp-_ pf 4- -
④dt

G (A) -
GCPJ =
-

f F-(f)dt
p
GH
O
pf Feet
-
= -

t.co#J--pfFltdt
: .

Eg :
Solve
of sing du
Sol •
I
:

find] =p / ptgzdp =
tariff ;
Iz
'
tan coldp
-

p
=
-
=

Now ,
Cinna] Iz
time L = - tailp

%P
of
"

sing dn -

Iz -

•
lamp .

Put p = 0: And
of sinus
dn -

Iz
eg solve :L
@IgJ
:

Sol :
I @ sn]
pts Division by
= .
n means
,
integrating from Pto "

1T¥]
=p / •pPgdp
"
-
Igenlp ↳ D1 ;
I .

14¥] not defined

:
is
.
.

ftp.?- esJsoe:i-E: - msJ- if pY:÷i-¥÷]

-oT- iI: -.it-
eg : i

it
=
e- Ncos 2kt sink
G-
"
e-

eg
i[nÉn]
:

ftp.?z-pgf--f--ze-ksinhv-n
=

"
eg solve it
[ log ( Pp÷1) kno
:
=

1 [ fast =

log Ipos) -

hog CPD
l [ n fed [¥
→
É]
-

I [ refers]
¥-2
=

nf @ 2 t'
[ ÷zJ
=

= 2 Sinhrn

Solution DE
of using Laplace Transform -
:

eg solve '

y
:

y 0
-
=

Taking Laplace transform both sides :

on

L[
y]
'

hey] 0 -
=

p Icy] yds ily ) - -

- 0

(p 1) LEY] -

yes -

ily] ti -

:
y cen ( lap beef
taking
-
invase
-

on
eg
: "
y -
y
- O

Sol : '
portly) -

Pylos g -

+ Ky] - °

(Pats) hey] = UP tea

Legs
4£ ¥→ -
-

y =
Glossed ↳ sink

lg : "

ny *
ylang - 0 .
Given :

y @3--0 ,
'

y ① =L
Taking Laplace transform :

C-
Ddg ( partly) pylos y play] get
'
- -
+ -

+ c-
* HEY ] ) -0

C- D [ p2dgp[HyD -
aptly] -
l-
PHY) -14-13
Eep :*
d§¥÷ p÷dp -

i.lu/LEyJf-l-zlnfp2-I1tlnc--7LeyJ--cli+py
-

-112
 "'
step g- (- it
tpgj
G- f- ftp. E. ⇐ ⇐↳
=
-

i.
e. - - - -

]
lqt-cpt-iqii-il-z-3-z.pk - i - i -

)
y =
4s :& -

*
II. a
- -
- -

-1
Tom yd) =L
i.
y
1
= C .

To @ -1
y -
Tola)

Note :
I [ Told]

¥5s
-_

Convolution Theorem - :

fH*gH-_of pen tgltldt

"
-

t[fH*gH] - L
floe tgltdt]
-

=
L[ few] L[gcr☐ .
Proof - :

ifftfen -

tsgctdt ]
of
%Mµ"f( )dt]dn
-

) get
a- t

of ? / e- Prefer
"
-
-

tgctdtdn put n -
t-
y

=ÉPU→Ñf④g(
KtYflysdyJdl-LEfCnD-F@Jlr]CPJeg.i
g) dydt :[ Ptgµ[o
e-

[•÷* ] if Is
-

,
Solving integral equations using Laplace Transform :

f- H -
y@ qfkln-ty.lt)dt
is
y an unknown .

Taking laplace transform - :

L
[ fed] =
L[ y] & L[klnDL[ y]
→ L[g) = L[ ten]
It L[kens]

solve
cg
: :

yw - n' 1-
of Linen - t)
yctdt
Sd " both sides :

Taking
-

laplace on

ifsinn) icy]
L[ g)
Ipg
&
.

1cg ] -

Ipg →

h[pY¥
ily] to %
: . - -

g- signs
 BESSEL FUNCTION
Properties of the Bessel
function :
-

[ xp T.ph] XP Jp , Cx)
1)
%
=

2)
da [ x-P Jp (a)] -_ - x-P Jpn ( n)

3) Tp ' @ -
t.f.Jp.la) Jp→w] -
→ from 142

A) Tpa (a) JPY4 Tp in → recurrence

2dg
-
-

= relation

Orthogonal properly :
-

If Am and Xn are two zeros of Tp then ,

:

/ Tl Jpltmk Jp Ann) doe

{ & Jpñ
Q m *n
⇐
o
( tn) man
,

2h - P
Tp
"
=
I C- D ( a/ 2)
n ! T ( nap+1)
I Bessel 's
function of first kind
 LAPLACE TRANSFORM
Laplace transform - :

A
L [ few] -
of e- Pnf @ dn =
Flp)
Properties of Laplace Transform - :

1) I [ can] I 2) I [ rn]
= =
Until
p
¥
-
a
.

3) L Einar)
page a) tifcosan]
-

p÷zaz
=

5) Lffinh an] = a- b) I @sham]

p2_air
-_pPz-az
If L[fenD-_ FLPJ , then g) l[Rf@B= C- 1)
dip
1) I @ fed
am
=
Flp -
a)
5) L[
2) L[film] Pflp )
nappy] -1-15%4
- -

f@ 6)
t.GL#J--pfF-ltdeLflCnJf--p2F(pj-pf0J-fk0J
3)
L[ oftefltldt]
F÷
-
 LEGENDRE and GAMMA
's
Rodrigue
formula Pride
-

In ddn-nnflnt-s.IT
,
legendre
npn @ =/ 2n -
1) a Pn , @ -
(n 1) Pn -64
-

Orthogonality
e)
-
Hm @ Pnlnsdn
I
= 0

ii)
d
) picador = 2
/ (2n -11
Gamma
function - :
•

TIM __
of e-
✗
sent dn
Cry -_

of Kum da ,
n > 0

Tcm) Mn)
7M¥
=
=
@ 1) In -1
TH
-

, - I

515 ) -
4 ! ;
54g) -51T 9ns __ •
Laplace transform of some important functions -
:

Unit step Ualt)

function -
:

I - •

{
° tea
* valet
.
=

I t za la t
,

[Walt] of 1-
I

Ptuatydt-fe-ptdt-e-aP.pl#if lt-a1UaltD-(afFlt- )
__

a) Eptdt ut
t
y
- a -
es

e-Ply
- a)
f- (g)
dy - e- Flp)
i. [
'
[e- 9
Ffp;] =
f- Lt -

a) Walt

[ FIT]
'

eg : solve e-

Sol : Here Ffpy 3-

-_¥y
a -
, ;
,

it
[ ¥1 ] since 3) uit
i. -
-
 unit
Dirac delta or
impulse function - :

{
too
I.
felt -
{ ,
a
e- b- I ate

0
,
t > ate

{
0 t -4 a
salt - him
feet -
,

E.→ o as t-_ a
,

then :
Tf g is
any
continuous function ,

es

of g Salt dt =
glas →
filtering
property
proof :

fgftdt-1-afgltdt-gltofgt-aaff.edu]
at E

of Ttt
-
few
where a 2 to Late

In the limit E- 0
,

him felt = fact ,

then
Ed 0

of °g⑧8dtdt- goes [ as a
a)
< to <
 Pt
If we put g⑨-_ e- ,
then

tYlt8aHdt= L[ Salt] - e- 9

For a- o
, L [811-1]--1
eg :
f @y =
{ Sinn
0
,
,
0 In
or > IT
ITT

Lf f- a] =

of %Ff.cn dn ,
"
=
of opmsinndn 40

I
: .

[ try] = Item
T
G :
f- End =
[ n] - m ,
m In cmtl

I [ fed] ←
of %M[ a) du

ftp.nodnqf?etndn-fbeiP?z.dn- - - -1-pLeiP-e-2P-
-
•

e-Me . .
. - .
)

I [ as] -

Lp ( Eder ) -

pEe÷D
En :
f-Cry is a
periodic function of period a

i.e.
, f- In +a) =
f-(a)

%Ffcydn-ofaetfendnqfetmfwdnna.tn
2A
I[
frog -_
of
-Da
f

÷ [email protected]
- . - -
.

-
ut n -
gotra
,

"

nÉ of e- PIM
'
flntnaldy
-
nÉ (e-4)If "etYf(ydy=

i.
Laplace function of periodic function a

with period is
given by
a :

[ tens]
I. ofaetitfydy
I -_
Eg of %;;÷dt
: Show that
Izék
=

Sol : Take Laplace loans form - :

iH%;¥- ] :/ I:P;;÷dtdnat

If ?÷µ 1%9 "

cosntdndt
↳ integrating by poets
•

of ?÷tap%ilt Fp:
ftp.I-ei#e-sfdl--,PqIz(-p-4--Izip'-s
⇐ -

laplace
taking
,
Now , inverse

• [%Y→dt -
Iz e-
n

solve Con Dy Kn&9) y (8%-0)

"
eg : '
my & o
-
-
- -

Taking Laplace ltans

form on both sides,
EH
dgpfp-y-pyloj-yYOJ-%dEPY-yof-f.pe y -

AED
of -94=0
-

L[ y ] =p 9-
'

you ; if gig =
pay pylos 8D
- -

L@
f AT C-D
%
-

( p 1) ( Pti)
G- 31 pt&4-
-

+ 0

% & 3%-1--0
ln Y -13 lnlp -
1) - Inc

Y
¥7,3
: . -

y -
green
 Fourier series

f- (a) = § an Cn - no )n
n = 0

f- ( ko)
"
where an =

ni .

If we are
given a
function periodic on -4,17]

I
fanwsnntbnsinnn1f@J_Ao_zn-1AoiAn.bn
+

are called Fourier coefficients

t
IT
I.

-TT
/ costndn - 0
2 .

f sin nndn __ O

-
M
IT

3-

-17
/ cosmnwsnndn =

{
°

IT,
, Mtn

m - n

/ sinmnsinnndn
{
° Mtn
4. ⇐
.

-17 IT, n
m -

5. Sinmncosnndn - O
 f- ☒ É lanwsnn & bnsinn)
agg
=
&
n-_ 1

Integrating from
t b- IT
-

" IT

→
/ f- @ dn =

ag-fz.dn-ao.tl
-1T

IT
ao -

1¥ f- (a)dn

To obtain an ,
and
we
multiply the
function by wsn "

integrate
t
from -17 to 1T :

TT
-
/ fer wsnndn
) = Than

t
an
,&f@
-
wsnmdn

Similarly , I

bn =

4- →
fttrsinnndn
eg :
fly -_ n -1T In c- IT

"
Sol Ao
↳
: -
ndn - 0
Ah =

¥ fttkcosnndn = 0

-17
Tl

fnsinnndn-fnfwsnn-n-11-sinnn.IR?,-Tli.bn--2(-
on =
Ig
MI
2)
I
n -41

i. n = É -
2 C- 2)
Sinn
h
n
-
z

on
, n = 2
( Sinn 1-zsin2n-l-zs.in3m )
-

Note : While
solving integrals
series.
in
fourier
cg :
fnsinndm →
keep differentiating
the

I 1-
first function
and integrating
the second
.
Add the tetons
reach 0
turbot you
.

fnsinnndn-nl-cosn-nnl-IL-sin.LI)
frihsinnndn = n'
f-WII) 2nf-sinnnn-fd2.com
-

'

F
 f- ☒ É lanwsnn & bnsinn)
agg
=
&
n -_ 1

The euists when the series converges

equality only
.

ale
The condition for convergence
:
-

Dirichlet 's conditions - :

try and bounded [-1-1,17] with finite

periodic
on
is
and / minima
discontinuity
morning
of points of
.

no .

of flu converges
series to
Then the fourier

12 [ fln-seflntif-l-zff-ln-fryf.tn

{
0 → In 20
G find
:
the
fourier feng ,
,

0 2- KITT
of
series K
,

¥ [1-4^-2]
Sol : ao -_

Ig ,
an -

bn = C-2) not
a-

§,gEb If
"
cosnn
fins
:
-

g- *
-

sinning
"
•
 that
Now , this
fourier
series , prone
using
TGI = I *
Igg Igg
& & . - . . -

Now, lets choose n - TT

At n - TT

124-1177 -
flits] -

tf
+
§%¥nH -17^-11-25

& ( THO) 1- ¥11 Ig

-
stat )
- -
.

- -
.
-

TGI It
1g +1g
'

.
& . . - . . . .

=
-

[email protected]
¥
ao -

"

/ fry sinnndn

Now ,
if f- Cry is even :
fl -
od -
fry
T

⇐ ofFey dn -2g ffcnscosndn

ao -
; an
.

bn = 0 .
Similarly , if f-(a) is odd !
f- C- a) = -

ffn)
do - 0 ,
an -_ 0 : bn -

¥ ,
[f-(a) sinnndn .

N - 2
( Sinn Ig Sinan & )

{
-
. - -
. .

" c- ETTTD
"
Iz ¥ / went tggwszn
=

,
-

, . . .

when R C- [ 0,17J ; then n - INI ,

then we

on cosine
sine
represent functions
as
can either of the

series .

eg : find half range sine series of

f-H - cos n
,
n c- ]
Eat
TT
bn
¥ / flnysinnndn [a. and Of
=
an are

%-[÷÷]
-

,
n -1-1

but
,
2-qofthosnsiandn-0.es
by -

cos n - E bnsinnn
n =L
=
54-[1%-11]
n= 2
G find cosine series of :
half range
: -

fed cos = n

{
° -1-1
Tosu
n

¥/
'
Ah =
cos nndn =

, TT
, n =L

Ao = 0

! Cos u
.
=
Az cos n = Cos n

Fourier series
for Arbitrary interval - :

suppose that there is an

arbitrary interval
E- 1 ☐ ,

Put t [-171-1]
ITI EL ☐ t c-
-
, n e- ,

f- * =
f- ( 1¥) -
get / gets -
agingzlancosntobnsinnt
f- Cry E En cosCtd) -
[email protected]
i.
=
Ag &

3g § Ig
-
find dn , an -

nscosfntnjdn
bn -

If - ftfln sin
(In_)dr
lg : find
fourier
series of the
function

{
0 -2 ↳ neo
try
,
=
1 0 ↳ v42
,

Ao -

& tfcndn -

Ig of 3. dn - 1

bn
Ing [1-1-17]
an = 0, -_
 Systems of First Order
Differential Equations

{ dy-dt-asn-bzy.gs
•

{
am As
__
+
It n - nlt)

ylt
1 .

If n -
ng and y -
y are solutions then ,
A-
Nz y -
J2
N =
4Mt Crm
Y -
9h +
ay, } → General Sol

f
2. w N

)
- m
- 0 oh non
zero
ly yn
Is
How to solve - :

Let D=
dat .
Then ,
we have

① → (D- a) n Put Aemt

big n -
-
- o

② → ask Bernt
+
brfy o
y
-
-
- -_

i.
① → Cm -
9) A -
b, B - O

② → -
ar A &
( m - b
2) B - O
The determinant of these eg.es should
be 0 .

i. M -
ar - b,
- 0
→ MI
-
As m - bz → Mz

for m -
mi A Ai B- Bl
=
, ,

for m -
mz , A -_
Az B- B2
,

if i .

✗ = A , em 't Az @
Mat
,

y = B, emit Bzemrt
,

iif when my Mz - Mto

-

{
A*em*t
{
(AHAztem*t
n -
"-

Y -
B*em*t Y -
( Bi&Bztfem*t
iii)
my -_
auditor , mz -
ai ibi
de
-

A- Artittz &B-_ Biti Bz

N-
(Ai+iAz)ea&iblt

eca-iblt.in
F- ( Biting
eat -
( A ,wsbt -

Azsinbt)

y -_ eat / B, cosbt -

Bzsinbt
eg
dn-dt-k-Y.dz
:

hn
-
2g
-

Sol :
write in operator form
① -
in -

g
- o
; -
hn - (D-2) y - O
 Non
homogeneous system - :

dn-dt-arn-biy-lf.lt)
{
" ←
" " → "" * Kp

%T = as " →
bzy +
f.µ y -
aye
cry~&Yp
Method
of variation of parameters - :

Toke n -
V1 Mt Vzrkr ;
y -
Viji & V242
substituting these
eqs
into the
system ,
we
get - :

vina vina t.lt) :

-
viy , - wife -
fit
vis
f ti
fl na

ya/
M fl
; v1 =

H tr

/ gal
M M

a
Im
m

YI Ya
Variable Elimination
approach -
:

eg diff
:
51-+2 ;
nay diff hn
-
-
-
-2g -81--8
*
oh, LD 1)
-
n -

y - -51--12, - hn +
Indy - -81--8

operating (13+2) on ② .
( D= dat, and

simplifying ,
we
get
↳D-6) se - -181--9
?⃝
i. m2 & M 6--0 → m =-3 2
-

ezeftnp-i-l-ist-91-%1F.ge/fIgt)(D2&D-
Kh -
G e- 3T →

G)
Up
% ( 1- Bye ) (-181--9)--31-42
=
. - .

i. Bt
n
q e- wettest 42
-
→

Now, to
use n in the
original equations
find g.

Converting higher DE into a linear

system of DE - :

Eg doggy
:
+
g- 0
dd%
{
- s
Let system of
defenses
r,
y -
.

e- lineal
In
-
=

equations