PROJECT ON:

o FOURIER SERIES AND FOURIER INTEGRAL

SUMITTED
BY

DILIP PAUDEL

PURO8OBCTO27

KATHMANDU 8NEPAL

To

MR.GOVINDA NEUPANE
TRIBHUVAN UNIVERSITY
INSTITUTE OF ENGINEERING

DEPARTMENT OF ELECTRONICS AND COMPUTER ENGINEERING

PURWANCHAL CAMPUS
DHARAN, NEPAL
 The primary purpose of this Report is to consider some of che

most important concepts of Fourier series and Fourier integrals.

Most of the períiodic funct ions discussed in this Report are defined

and integrable over the interval -, T and the analysis of the

Pourier seríes is carríed out for funct ions on this interval; however,

it is shown that thís can be extended to funct ions defined and in

tegrable over more general intervals such as, -L,L]. The analysis

consists not only of consLder ing condítions under which a periodic

function can be represented by its Fourier series, but also of con

8idering condítions under which a Fourier series may be different lated

and integrated term by term, Conditions are also given for which a

nonperiodic function can be represented by its Four ier integral.

When the Pourier seríes of a funct ion is being studied, che

integrands giving rise to the Fourier coefficients of the function

are assumed to be integrable in the sense of Riemann. The extension

of RiemAnn integration to the case of improper integrals is made when

necessary. This is the case for the Four ier int egral of a function,

Four ier series and Pourier integrals are interesting in them

selves and as a valuable tool in solving physical problems . In this

Report var ious problems are discussed, and the applicat ion of Pourier

analysis to them is shown.

 Letter of approval

This project work entitled "

FOURIER SERIES AND FOURIER
INTEGRAL" submitted by Mr DILIP PAUDEL. For the partial
fulfillment was carried out under my guidance in the academic
year 2025 .Mr DILIP PAUDEL had performed the project work
sincerely and satisfactory

Mr
 Acknowledgement

Acknowledgment

Iwould like to express my sincere gratitude to Mr. Govinda

Neupane, our respected Mathematics teacher,for his invaluable
guidance and support throughout this project. His encouragement
and insightful feedback have been instrumental in the successful
completion ofthis work.

Iam also thankful to Purvanchal Campus forproviding the

necessary resources and a conducive learning environment. Lastly, I
appreciate the support of my peers and mentors who have
contributed directly or indirectly to this project.

Submitted by:

Dilip Paudel

s......

25 feb, 2025
 Table of Contents
1. Introduction

2. Fourier Series

a.Definition and Concept

b.Derivation

3. Fourier Integral

a.Definition and Concept

b.Derivation

4.Applications of fourierseries and fourier integral

5. Conclusion

6. References
 INTRODUCTION

Attention was drawn to trigonometric series in the middle of the

eighteenth century when various ma themat icians were studying the vi

brations of stretched strings. The mathematical theory of such vi.

brations comes down to the problem of solving the partial di£ferent ial

equation u =b u The general forn of a trigonometric

series, involving both sines and cosines, arises natural ly in con

nect ion with another physical problem, that of conduct ion of heat.

It was in the course of hís study of this problem that J. B, Fourier

announced in 1807 that an arbitrary funct Lon £ (x) could be represented

in the form

(a, cos nx +b, sin nx)

hal

with an &nd bn given by the integrals

cos nx dx, ne0,1,2, ...

bn sin nx dx, nl,2, *.

n
The trigonometric series given above is called the Four ier series

corresponding to £(×) and the constants an and bn are called the

Fourier coefficiente of the function,

We should point out ,however, some trigonometr ic series are not

Pourier series. A trigonometric series of the form

2+ cos nx +bn 8in nx)

is not called a Fourier series unless there is some function f (x) from

which the coefficients in the series may be calculated by the formulas

given above for an and bn

The integrals defining an and ba must be meaningful, i,e., the

integrals exist either as proper Riemann integrals, Lebesgue integrals

or as improper integrals. Throughout this Report when the Fourier

series of a function is being studied the function will be assumed

Riemann integrable. Most of the functions arising in ordinary ap

plications are bounded and are continuous except possibly at a finite

number of points and are therefore integrable in che sense of Riemann.

Four ier series are useful in dealing with periodic functions

but if the function is not periodic, it cannot be represented by a

Pourier series. There is, however, an expression involving improper

integrals instead of infiníte series, which under suitable conditions

gives a valid represent ation of a nonperiodic function, Then we say

the function is represented by its Pourier integral.

The purpose of this Report is to acqua int the reader with some

of the most important aspects of the theory and applications of Fourier

series and Fourier integrals .

 Fourier Series of a Functíon

If £(x) is a function which is de fined on the

Definition 1.1.

interval asx sb and such that the Lntegrala

(1) cos nx dx, n0,l,2, .

(2) sin nx dx, nl,2, .
have a meaning, we form the serles

Bo (an cos nx + b, sin nx)

2

and cal1 thís the Fourier series corresponding to f (x).

The funct Lon determines its Fourier serLes quite apart from any

consideration as to whether the serles converges or, if it does con

or not its sum is (x). If the Fourier series of a

verge, whether f

Ln advance whether it con

function £ (x) Ls forned without deciding

verge s to f(%), then we write

£(«) o (a, cos nx bn 8in nx).

2

This notation means merely that the Fourier series on the right cor

f(x). The sign can be replaced by the

responds to the functíon 2

only if we succeed in proving that the series converges and

sign

that its sum equals f (x).

Before we find the Fourier series correspond íng to a particular

Éunct ion, we should give an explanat ion of how we arrived at the in

tegrals for the Fourier coe fficients an and bn given by (1) and (2).
For
First consider the orthogonality of sine and cosine funct i ons .
any int eger n >0

(3) cos nx dx 0

(4) sin nx dx =0

(5) cog2 nx dx =

(6) sin nx dx =T
Let m and n denote positive Lntegers. Then

(7) kos nx cos ux dx 0

1f mn

(8) sin nx 8in mx dx 0

1f mfn

(9) sin nx cos mx dx0 for any m and n.

Suppose the function f (x) of period 2 T has the expansion

(10) cos kx +by sin kx).

It will be assumed that the series (10) and the series to be wrítten

later can be integrated tern by term. Integrating (l0) from -n to T,

we obtain
 cos kx dx + bx sin

k:l

By (3) and (4) all the integrals in the sum vanish, so that

dx =T ag

Next, multiply both sides of (10) by cox nx and integrate the result

from - to n, as before, obtaining

(11) cos nx dx o
cos nx dx cos kx cos nxdx
2

sin kx cox nx dx

By (3) the first integral on the right vanishes. By (7) and (9) all

the Lntegrals in the sum vanish except one. Thus (11) reduces to

cos nx dx = an cos nx dx.

Now by (5),

(12) Bn ()cos nx dx, n0,1,2, *.

Similar ly,

(13) sin nx dx, n l,2, ..

 Before computing the Pourier series corresponding to a particular

function £(%), there is a considerat i on which will sonetimes simp lify the

calculation of the Fourier coefficient s.

Definit ion 1.2. The funct ion f(x) is said to be an even function

if f(«) =f(-x) for every x.

This definition implies that the graph of any even function is syunetric

with respect to the y-axis. It follows from the interpretation of the

integral as an area that for even functions defined and integrable on

the interval ,n

(14)

Definition l.3. The function f(x) is said to be odd if

£(-*) =-f(x) for every x.

The graph of any odd function fE (3) is syunetric with respect to

the origin. Thus, for odd functions defined and integrable on

che interval

(15) dx =0.
 Let f(x) be an even function defined on the interval -n,n
Since cos nx (n=0, 1, 2,...) is an even function, then by Property

(1) the function f(x) cos nx is also even. On the other hand, the

function sin nx (n=l, 2, ...) is odd, so that the function f(x)

sin nx is also odd, by Property (Li). Then, usíng (12), (13), (14) and

(15), we find chat che Fourier coefficients of the even functíon £(x)

are

(16) £(*) cos nx dx=2 f(«) cos nx dx, n=0,1,2 *.

(17) sin nx dx =0, n=l,2, .

Therefore, the Fouríer series of an even function contains only

cosines, i.e.

an coS nx.
The series above is known as the Fourier cosine series.

Nw let f(%) be an odd funct ion defined on the interval-, n.

Since cos nx (n=0, 1, 2, ...) is an even function, then by Property

(11) £«) cos nx is an odd function. On the other hand, che function

sin nx (n =l,2, ...) is odd and by Property (1) the function f (x) sin nx

is even. Then, using (12), (13), (14) and (15) we find that the

Fourier coefficients of the odd function £(x) are

cos nx dx =0, n=0, 1, 2, .

sin nx f(«) sin nx dx, n=l,2, ....

dx z
-n
Therefore, the Fourier series of an odd function contains only sines,

i.e.,

£ (×) ) bn sin nx.

This series becomes the Fourier since seríes.

 The Four ler Integral

Pourier series are useful in dealing with periodic functions.

If a functíon ís defined for all real values of x, but is not periodic,

it cannot be represented by a Fourier series. There is, however, a

method involving improper integrals instead of infinite series, whích

under suitable conditions gives a valid representation of a function

£(«).

Definítion 4.1. The expression

(1) l/T du
(e) cos [u(e-] dt

is called the Fourier integr al of f(«).

The analogy between a Fourier integral and a Fourier series may

be seen as follows: For any uz0 let

(2) a (u) =1/T F(t) cos ut dt

b (u) = l/T F(t) sin ut dt.

The funct ions a(u) and b (u) are ana logous to the Fourier coefficients

an and bn. Then forn the integral

(3) cos ux + b(u) sin ux du,

which Ls analogous to the Four ier series (10).

Now,

a (u) cos ux +b(u) sin ux l/T E(e)cos ut cos ux

+sin ut sin

Thus, the integral (3) can be put in the form (1). A1l these consi

derations assume that the integrals in question are convergent.

The fol lowing results for the Four íer integral are analogous

to those for the Fourier series of a funct Lon. If £(*) is an even

function, £(c) cos ut is even and f(t) sin ut is odd, so that b(u) =0 and

a (u) =
2/T cos ut dt.

In thí case the Four ier integral takes the form

2/T cOs ux du £ (t) cO8 ut dt,

which we call the Fourier cos ine integral. For an odd funct ion
f (x), we can show that a =0
(u) and

b (u) = 2/1 sin ux du sin ut dt,

we call this the Fourier sine integral.

The conditions given in the following theorem, called the

Four ler integral theorem, Lnsure that £ (*) may be represented by its

Four Ler integral.

Theorem 4ol. Let f(«) be sectionally continuous in every finite

interval (a,b), and let converge. Then at every point x,

(-o0< x<), where f (x) has a right-and left-hand der ivative, £(*)

is represented by its Pourier integral as follows:

Proof: In every interval (a,b), £(«) 8atisfies the conditions

of Theorem 2.3, so that

at any point x (a <x<b),where f (*) has a right-and left-hand

derivative.

To prove (4) we have to show that if (0,

(6

for all sufficient ly large values of .

 Applicat ion of Fourier Series and Integrals

In this section we shall discuss how Four ier series and integrals

may enter into the solution of certain boundary value problems. First we

will consider the vibrating string problem.

The equation of motion of the vibrating string fastened at both

ends is

(1)
Our problem is to find a solution y(x, t) that satisfies the equation

of motion and boundary conditions,

(2) y (0,t) = 0,y(L,t)=0

(3) dy(x,0)/dt=0

(4) y (<,0) =f(*).

First we shall try to find solutíons of (1) which satisfy the

boundary conditions (2) and (3). We shall use the separation of

variables method to find the solut ions.

This method consists in looking for solut Lons of the form

(5) y («,t) XT

where X is a funct ion of x alone and T is a function of t alone.

According to equat ion (5),

dx

dt2

If these expressions are substituted into equation (1), we get

If ve divide through by a xT, then equation (1) can be

rewritten as
 = dx2

Since the member on the right is a function of x alone, it cannot

vary with t; it is equal to a function of t alone, however, and thus

it cannot vary with x. Hence both members must be equal to a constant ,

say -w, so that

dx2

de

(6) dx + w2x =0
dx2

(7)
de2

It may not be obv ious to the reader why we chose -w instead of a

constant B. Suppose that we had chosen a constant B so that

X"- BX =0
T" - Ba2T = 0.

The general solution of the first differential equation i8

where M and M are arbitrary constants. If B is positive there are

no values of M1 and M2 for which this function X satisfies the con

ditions, Therefore, we choose the const ant -.

The general solution of equation (6) can be wrítten as

XAs in wx +Bcos Wx,

where A and B are arbitrary constant s.

If X(0) =0,the constant B must vanish. Then A must be different

from zero, since we are not interested in the trívial solution X(*) =0.
So 1£ x(L)=0, we must have

sin wL 0.

Hence there is a set of values of w, namely,

Wn T/L

for which the system consísting of equation (6) and b oundary condit ions

(2) has solutíons. These solutions are of the form

XE Asin n mx/L.

Equation (7) has as its general so lut ion

TCsin hat + Dcos wat,

If it is to satisfy condition (3), then T' (0) = 0. Thus the coefficient

C must be zero, so we find that

T= Dcos wat.
Now if we substitute the value n T/L for w, we have

T= Dcos nyat/L,

Therefore all the functions

A,sin n mx/Lcos nyat/L, n=l, 2, ...

are solut ions of our part ial dLfferent ial equation (1).

Any finíte linear combinat ion of the solut íons will also be a

solut ion of (1) and satisfy condit ions (2) and (3). When t =0, it wil1

reduce to a linear combination of che functions

8in n x/L.

Thus cond ícíon (4) wi1l not be satisfied unless the given function

f («) can be written as a linear combination of sine functions.

Consider the infinite series,

y=) Asin (n x/L) cos (0 at/L).

This series sat isfies equat ion (1) provided it converges and it tera

wise different iab le; it also satisfies conditions (2) and (3). It wil1

Batisfy the nonhomogeneous condition (4) provided the numbers An can

be determíned so that

P(*) = )Ansin n nx/L,

This series is the Four ier sine series for £ («), so

Anl/L £ (*) sin nx/Ldx = 2/L F(«) sin n nx/Ldx.

Thus the solution of equat ion (1) subject to the boundary conditions

(2), (3) and (4) is

y(x,t) = 2/L) (sin nmx/Lcos n mat/L F(«) sin n x/Ldx) .

From the nature of the physical problem, we expect that there

should be just one solution y (*,t) of equatíon (1) which satisfies

the boundary condit ions (2), (3) and (4), and this solution will

represent the equation of mot ion of the string.

Another boundary value problem which we can solve with the aid

of Four ier series is that of finding a formula for the temperatures

in a slab of homogeneous material with the faces of the alab kept at

temperature zero.

In thi8 problem the temperature is a functí on of the variab les

x and t. At cach interior point thís function u(x,t) must satisfy

the heat equati on of one-dimensional flow,

(1)
dy2
0< x< Wt >0.

It must also satisfy the boundary condíeions

(2) u(+0,t) =0, u(T-0,t) = 0

(3) u(x, +0)=f (*).

To find a particular solution of (1), we again will use the

separayíon of variables method, First find a solution which

satisfies condit ions (2). Let uX(«)T (t). Hence ,

 Conclusion
The Fourier Series and Fourier Integral are fundamental

mathematical toolsfor representing and analyzing

functions in both periodic and non-periodic domains.

Their applications span multiple disciplines, making them

essentialin engineering and scientific research.

References

1. Kreyszig,E. (2011). Advanced Engineering Mathematics. Wiley.

2. Oppenheim,A. V., & Schafer, R. W. (1999). Discrete-Time Signal

Processing. Prentice Hall.

3. Bracewell, R. (2000). The Fourier Transform and Its Applications.

MCGraw-Hill.
 BIBLIOGRAPHY

1. Byerly,William E.
Fourier Series, Boston:
Ginn & Comp any, 1895.

2. Carslaw, H. S.
Theory of Fouríer's Series and Integrals, London:
Macmillan and Company, 1921,

3. Churchill, Ruel V.
Four ier Series and Boundary Value Problems, New York:
McGraw-Hill Book ,
Comp any 1941.

4. Hardy, G. H., and W. W. Rogosinski.

Fourier Series, New York:
The Macmillan Company, 1944.

5. Jackson, Dunham,
Fourier Series and Orthogonal Polynomials, Mena sha, Wiscons in:
George Bantan Publishing Company, 1941.

6. Symon, Keith R.
Hechanics, London:
Addison-Wes ley Publishing Company, 1960.

7. Tít chmarsh, E. C.
Theory of Fouríer Integrals, Oxford:
Oxford Uníversity Press, 1937.

8. Tolstov, Georgi P.
Fourier Series, Englewood Clíffs, New Jersey:
Prentíce-Hall, Inc., 1962.