Topic 06: Solving Nonhomogeneous ODEs

Second Part of MA102 Mathematics II: Winter Semester of AY 2024-2025

MGPP, SN, SHB: IIT Guwahati

Sections 4.3 to 4.4

of
S. L. Ross, Differential Equations, 3rd Edition, Wiley
Learning Outcome

I Method of Undetermined Coefficients

I Annihilator Method (Method of Undetermined Coefficients by Annihilator
Approach)
I Operator Method
I Method of Variation of Parameters
 Method of Undetermined Coefficients
for
Nonhomogeneous ODEs with Constant
Coefficients
Nonhomogeneous term is Polynomials, Exponential Functions, Sine &
Cosine Functions
Method of Undetermined Coefficients
Consider the nonhomogeneous differential equation

dn y dn−1 y dy
a0 n + a1 n−1 + · · · + an−1 + an y = F(x), (1)
dx dx dx
where the coefficients a0 , 0, a1 , . . ., an are real constants and the nonhomogeneous
term F(x) is a function of x (It will be restricted to some specific forms).
The general solution of (1) is given by

y = yc + y p , (2)

where yc is the complementary function, that is, the general solution of the
corresponding homogeneous equation

dn y dn−1 y dy
a0 n + a1 n−1 + · · · + an−1 + an y = 0, (3)
dx dx dx
and y p is a particular solution of (1), that is, any solution of (1) containing no arbitrary
constants.
Method of Undetermined Coefficients

Idea: Make an educated guess of the form of particular solution y p from the form of
nonhomogeneous term F(x) and then determine its coefficients.

The nonhomogeneous term F(x) should be one of the following forms:

I Polynomials in x.
I Exponential Functions eax .
I Sine or Cosine Functions sin(βx) or cos(βx).
or finite sums or product of these functions.

Special Nature of These Functions: The derivatives of these Functions are

I Either itself
I or Constant Multiple of it
I or else Linear Combination of them
Guessing the Form of y p

The assumed form of the particular solution y p should be a linear combination of all
linearly independent functions that are generated by repeated differentiation of the
nonhomogeneous term F(x).

Rules
I No function in the assumed particular solution y p is a part of the complementary
function yc (Solution of the associated homogeneous differential equation).
I If assumed particular solution y p contains any terms that duplicate terms in the
complementary function yc , then such terms of y p must be multiplied by xk , where
k is the smallest positive integer that eliminates the duplication.

That is, the positive integer k is chosen to be the smallest integer so that no term in y p
is a solution to L(y) = 0.
In case F(x) is a polynomial Pk (x) of degree k, if some terms of Pk (x) are zero/missing,
then also the assumed form of y p must include all terms of a polynomial of degree k.
Case-1: Nonhomogeneous Term F(x) is a Polynomial Pk (x)
Consider
dn y dn−1 y dy
L(y) = a0 n + a1 n−1 + · · · + an−1 + an y = Pk (x),
dx dx dx
where the coefficients a0 , 0, a1 , . . ., an are real constants and the nonhomogeneous
term Pk (x) is a polynomial of degree k.
Assume that the complementary function does not contain the polynomial terms.

We assume the form of the particular solution y p as

y p (x) = A0 + A1 x + · · · + Ak xk .

Note: If some terms of the polynomial Pk (x) are zero/missing, then also the assumed
form of y p must include all terms of a polynomial of degree k.
Now, L(y p ) = Pk (x) yields a system of (k + 1) equations in (k + 1) unknowns A0 , A1 , . . .,
Ak that will be determined by solving it.
Example: Nonhomogeneous Term F(x) is a Polynomial Pk (x)

Find the particular solution y p of y00 + 3y0 + 2y = 3x + 1. Also write its general solution.
Step 1: Finding the Complimentary Function yc
Consider the associated homogeneous equation

y00 + 3y0 + 2y = 0.

Then, solving the auxiliary equation r2 + 3r + 2 = 0, we get r = −1 and r = −2.

Therefore, the complementary function yc is given by

yc (x) = c1 e−x + c2 e−2x ,

where c1 and c2 are arbitrary real constants.

Continuation of Previous Slide

Step 2: Guessing and Writing the Form of Assumed Particular Solution y p

Since the nonhomogeneous term is a polynomial degree 1, we assume the form of the
particular solution as
y p (x) = A0 + A1 x.
Observe that the functions in the assumed particular solution y p is NOT a solution of
the associated homogeneous equation (That is, not a part of yc (x)).
Step 3: Determining the coefficients of terms in y p from L(y p ) = F

L(y p ) = y00p + 3y0p + 2y p = 3x + 1

=⇒ 3A1 + 2(A0 + A1 x) = 3x + 1
=⇒ 2A1 = 3 and 3A1 + 2A0 = 1
3 −7
=⇒ A1 = and A0 =
2 4
Continuation of Previous Slide

Step 4: Writing the Particular Solution y p

The particular solution of the given ODE is

7 3
y p (x) = − + x.
4 2
Step 5: Writing the General Solution y
The general solution of the given ODE is

y(x) = yc (x) + y p (x)

3x 7
y(x) = c1 e −x
+ c2 e −2x
+ − ,
2 4
where c1 and c2 are arbitrary real constants.
Case-2: Nonhomogeneous Term F(x) is an Exponential Function aeαx

Consider
dn y dn−1 y dy
L(y) = a0 n + a1 n−1 + · · · + an−1 + an y = a eαx ,
dx dx dx
where the coefficients a0 , 0, a1 , . . ., an are real constants. Also a and α are real
constants.
Assume that the complementary function does not contain the terms that are scalar
multiples of eαx .

We assume the form of the particular solution y p as

y p (x) = A eαx .

Now, L(y p ) = F yields an equation in the unknown A that will be determined by solving
it.
Example: Nonhomogeneous Term F(x) is an Exponential Function aeαx

Find the particular solution y p of y00 + 3y0 + 2y = e3x . Also write its general solution.
Step 1: Finding the Complimentary Function yc
Consider the associated homogeneous equation

y00 + 3y0 + 2y = 0.

Then, solving the auxiliary equation r2 + 3r + 2 = 0, we get r = −1 and r = −2.

Therefore, the complementary function yc is given by

yc (x) = c1 e−x + c2 e−2x ,

where c1 and c2 are arbitrary real constants.

Continuation of Previous Slide

Step 2: Guessing and Writing the Form of Assumed Particular Solution y p

Since the nonhomogeneous term is an exponential function e3x , we assume the form
of the particular solution as
y p (x) = A e3x .
Observe that the functions in the assumed particular solution y p is NOT a solution of
the associated homogeneous equation (That is, not a part of yc (x)).
Step 3: Determining the coefficients of terms in y p from L(y p ) = F

L(y p ) = y00p + 3y0p + 2y p = e3x

=⇒ 9Ae3x + 3(3Ae3x ) + 2(Ae3x ) = e3x
=⇒ 20A = 1
1
=⇒ A=
20
Continuation of Previous Slide

Step 4: Writing the Particular Solution y p

The particular solution of the given ODE is

1 3x
y p (x) = e .
20
Step 5: Writing the General Solution y
The general solution of the given ODE is

y(x) = yc (x) + y p (x)

e3x
y(x) = c1 e −x
+ c2 e
−2x
+ ,
20
where c1 and c2 are arbitrary real constants.
Case-3: Nonhomogeneous Term F(x) is Sine or/and Cosine Functions

Consider
dn y dn−1 y dy
L(y) = a0 n + a1 n−1 + · · · + an−1 + an y = F(x),
dx dx dx
where the coefficients a0 , 0, a1 , . . ., an are real constants and F(x) = a sin(βx) or
F(x) = a cos(βx) or F(x) = a sin(βx) + b cos(βx). Here a, b and β are real constants.
Assume that the complementary function does not contain the terms that are scalar
multiples of cos(βx) and sin(βx).

We assume the form of the particular solution y p as

y p (x) = A sin(βx) + B cos(βx).

Now, L(y p ) = F yields an equation in the unknowns A and B that will be determined by
solving it.
Example: Nonhomogeneous Term F(x) is Sine or/and Cosine Functions

Find the particular solution y p of y00 − y0 − y = sin x. Also write its general solution.
Step 1: Finding the Complimentary Function yc
Consider the associated homogeneous equation

y00 − y0 − y = 0.
√
Then, solving
√ the auxiliary equation r2 − r − 1 = 0, we get r = (1 + 5)/2 and
r = (1 − 5)/2.
Therefore, the complementary function yc is given by
√ √
1+ 5 1− 5
yc (x) = c1 e 2 x + c2 e 2 x ,

where c1 and c2 are arbitrary real constants.

Continuation of Previous Slide
Step 2: Guessing and Writing the Form of Assumed Particular Solution y p
Since the nonhomogeneous term is sin x, we assume the form of the particular
solution as
y p (x) = A sin x + B cos x.
Observe that the functions in the assumed particular solution y p is NOT a solution of
the associated homogeneous equation (That is, not a part of yc (x)).
Step 3: Determining the coefficients of terms in y p from L(y p ) = F

L(y p ) = y00p − y0p − y p = sin x

=⇒ −A sin x − B cos x − (A cos x − B sin x) − (A sin x + B cos x) = sin x
=⇒ (−2A + B) sin x + (−A − 2B) cos x = sin x
=⇒ −2A + B = 1 and − A − 2B = 0
−2 1
=⇒ A= and B=
5 5
Continuation of Previous Slide

Step 4: Writing the Particular Solution y p

The particular solution of the given ODE is

(−2) 1
y p (x) = sin x + cos x.
5 5
Step 5: Writing the General Solution y
The general solution of the given ODE is

y(x) = yc (x) + y p (x)

√ √
1+ 5 1− 5 2 1
y(x) = c1 e 2 x + c2 e 2 x − sin x + cos x,
5 5
where c1 and c2 are arbitrary real constants.
Students to Workout in the Class
Solve y00 − 2y0 − 3y = 4x − 5 + 6xe2x by finding a particular solution using the method of
undetermined coefficients.
I Complementary Function: The general solution of the associated homogeneous
equation y00 − 2y0 − 3y = 0 is
yc (x) = c1 e−x + c2 e3x .
I Since non-homogeneous term is of the form P(x) + Q(x)e2x , where P(x) is a
polynomial of degree 1 and Q(x) is a polynomial of degree 1, a guess for a
particular solution is
y p (x) = (Ax + B) + (Cx + D) e2x .
Observe that the functions in the assumed trial particular solution is NOT a
solution of the associated homogeneous equation (That is, not a part of yc (x)).
I
−4 23 −4
A= , B= , C = −2, D= .
3 9 3
I The general solution of given non-homogeneous ODE is
!
4x 23 4
y(x) = yc (x) + y p (x) = c1 e−x + c2 e3x − + − 2x + e2x .
3 9 3
Method of Undetermined Coefficients: Form of y p (x)
Let pk (x) = b0 + b1 x + · · · + bk xk and Pk (x) = B0 + B1 x + · · · + Bk xk .
Let q` (x) = c0 + c1 x + · · · + c` x` and Q` (x) = C0 + C1 x + · · · + C` x` .
Let M = max{k, `}. Let P M and Q M denote polynomials of degree M .

Non-Homogeneous Term Form of y p (x)

pk (x) xk [Pk (x)]
aeαx xk [Aeαx ]
a sin(αx + β) + b cos(αx + β) xk A sin(αx + β) + B cos(αx + β)
 
pk (x) eαx xk [Pk (x) eαx ]
pk (x) sin(αx + β) + q` (x) cos(αx + β) xk P M (x) sin(αx + β) + Q M (x) cos(αx + β) ,
 
where M = max{k, `}
a eγx sin(αx + β) + b eγx cos(αx + β) xk A eγx sin(αx + β) + B eγx cos(αx + β)
 
pk (x) eγx sin(αx + β) + q` (x) eγx cos(αx + β) xk P M (x) eγx sin(αx + β)+

Q M (x) eγx cos(αx + β) , where M = max{k, `}


The non-negative integer k is chosen to be the smallest integer so that no term in y p is

a solution to L(y) = 0.
Method of Undetermined Coefficients: Form of y p (x)
Assume that NO function in the assumed particular solution is a part of the
complementary function.
Non-Homogeneous Term Form of y p (x)
k (Non-zero Constant)
5x + 7
3x2 − 2
x3 − x + 1
sin(4x)
cos(3x)
(9x − 2)e5x
x2 e5x
e3x sin(4x)
5x2 sin(4x)
xe3x cos(4x)
xe3x cos(4x) + x2 e3x sin(4x)
Method of Undetermined Coefficients: Form of y p (x)
Assume that NO function in the assumed particular solution is a part of the
complementary function.
Non-Homogeneous Term Form of y p (x)
k (Non-zero Constant) A
5x + 7 Ax + B
3x2 − 2 Ax2 + Bx + C
x3 − x + 1 Ax3 + Bx2 + Cx + D
sin(4x) A cos(4x) + B sin(4x)
cos(3x) A cos(3x) + B sin(3x)
(9x − 2)e5x (Ax + B)e5x
x2 e5x (Ax2 + Bx + C)e5x
e3x sin(4x) Ae3x cos(4x) + Be3x sin(4x)
5x2 sin(4x) (Ax2 + Bx + C) cos(4x) + (Dx2 + Ex + F) sin(4x)
xe3x cos(4x) (Ax + B)e3x cos(4x) + (Cx + D)e3x sin(4x)
xe3x cos(4x) + x2 e3x sin(4x) (Ax2 + Bx + C)e3x cos(4x) + (Dx2 + Ex + F)e3x sin(4x)
Example: Terms of assumed y p is a part of yc

Find the particular solution y p of y00 − y0 − 12y = e4x . Also write its general solution.
Step 1: Finding the Complimentary Function yc
Consider the associated homogeneous equation

y00 − y0 − 12y = 0.

Then, solving the auxiliary equation r2 − r − 12 = 0, we get r = 4 and r = −3.

Therefore, the complementary function yc is given by

yc (x) = c1 e4x + c2 e−3x ,

where c1 and c2 are arbitrary real constants.

Continuation of Previous Slide

Step 2: Guessing and Writing the Form of Assumed Particular Solution y p

Since the nonhomogeneous term is e4x , we assume the form of the particular solution
as
y p (x) = A e4x .
Important Observation: Observe that the term e4x in the assumed particular solution
y p is a solution of the associated homogeneous equation. That is, it is a part of yc (x).
So, we will find a smallest positive integer k such that xk e4x is not a solution of L(y) = 0.

Observe that if y(x) = x e4x , then L(y) = y00 − y0 − 12y = 7e4x , 0. So, we take k = 1.

The revised assumed form of the particular solution is

y p (x) = A x e4x .
Continuation of Previous Slide

Step 3: Determining the coefficients of terms in y p from L(y p ) = F

L(y p ) = y00p − y0p − 12y p = e4x

=⇒ (8Ae4x + 16Axe4x ) − (Ae4x + 4Axe4x ) − 12Axe4x = e4x
=⇒ 7Ae4x = e4x
1
=⇒ A=
7
Continuation of Previous Slide

Step 4: Writing the Particular Solution y p

The particular solution of the given ODE is

x e4x
y p (x) = .
7
Step 5: Writing the General Solution y
The general solution of the given ODE is

y(x) = yc (x) + y p (x)

x e4x
y(x) = c1 e + c2 e
4x −3x
+ ,
7
where c1 and c2 are arbitrary real constants.
Students to Workout in the Class

Using the method of undetermined coefficients, Find a particular solution of

y00 − 5y0 + 4y = 8e x .

Answer:
8 x
y p (x) = − x e .
3
Students to Workout in the Class
Using the method of undetermined coefficients, Find a particular solution of
y000 + y00 = x3 + x + 1.
The complementary function is yc (x) = c1 + c2 x + c3 e−x , where c1 , c2 and c3 are
arbitrary real constants.
Since the nonhomogeneous term F(x) = x3 + x + 1 is a polynomial of degree 3, we
guess the form of y p (x) as
y p (x) = Ax3 + Bx2 + Cx + E.
Observe that 1 and x are part of the complementary function yc (x).
Since L(1) = 0, L(x) = 0 and L(x2 ) , 0, we revise the form of y p (x) as
 
y p (x) = x Ax + Bx + Cx + E = Ax5 + Bx4 + Cx3 + Ex2 .
2 3 2

Then, the terms in y p (x) that are duplicating in yc (x) got removed. Now, we determine
these coefficients from L(y p ) = x3 + x + 1.

Final Answer:
x5 x4 7 x3
y p (x) = − + − 3x2 .
20 4 6
 Annihilator Method
(Method of Undetermined Coefficients by Annihilator Approach)
for
Nonhomogeneous ODEs with Constant
Coefficients
Nonhomogeneous term is Polynomials, Exponential Functions, Sine &
Cosine Functions

Refer:
Pages 141 to 149 of Differential Equations by Dennis G Zill, Cengage Learning, India
Edition.
Page 90 to 94 of An Introduction to Differential Equations by Earl A. Coddington, PHI
Learning Pvt. Ltd. New Delhi, 2011.
Annihilator Method
This method provides a procedure for finding a particular solution y p of
nonhomogeneous linear differential equation L(y) = F with constant coefficients,
where the nonhomogeneous term F is of specific form.
Idea: To solve L(y) = F , we find a linear differential operator M with constant
coefficients which makes M(F) = 0, that is, annihilates F(x). Consequently,
M (L(y)) = 0. It transforms the given nonhomogeneous equation into homogeneous
equation. The form of y p follows automatically once we have found an appropriate
linear differential operator M that annihilates the nonhomogeneous term F(x) in
L(y) = F .
Definition 1
If T is a linear differential operator with constant coefficients, and f : (a, b) ⊆ R → R is
a sufficiently differentiable function on (a, b) such that

T ( f (x)) = 0 for all x ∈ (a, b),

then T is said to be an annihilator of the function f .

Let D = d
dx and f (x) = e x for x ∈ R. Then T := (D − 1) annihilates f (x) on R.
Annihilator Method - Procedure
Consider the nonhomogeneous linear ODE with constant coefficients L(y) = F , where
F consists of finite sums and products of constants, polynomials, eαx , sin(βx) and
cos(βx).
I Step 1: Find the complementary function yc (x), that is, the general solution of
L(y) = 0.
I Step 2: Find the annihilator operator M that annihilates F(x). Then, operate M on
both sides of the nonhomogeneous equation L(y) = F .
I Step 3: Find the general solution Yc (x) of the higher order homogeneous
differential equation M (L(y)) = 0. Note that Yc (x) contains the terms of yc (x) and
some more terms.
I Step 4: Delete from the expression of Yc (x) all those terms that duplicated in yc (x).
Form a linear combination of the remaining terms and that will be the form of a
particular solution y p (x) of L(y) = F .
I Step 5: Substitute y p found in Step 4 into L(y) = F . Match the coefficients of the
various functions on each side of the equality, and solve the resulting system of
equations for the unknown coefficients of y p . Write the final expression of y p .
I Step 6: With the particular solution y p found in Step 5, form the general solution
y = yc + y p of the given nonhomogeneous differential equation L(y) = F .
Annihilator Operators

Non-Homogeneous Term F M = Annihilator of F

xn−1 , n ∈ N. Dn
eα x (D − α)
xn−1 eα x (D − α)n
sin(βx) or cos(βx) D 2 + β2
xn−1 sin(βx) or xn−1 cos(βx) (D2 + β2 )n
eαx sin(βx) or eαx cos(βx) D2 − 2αD + (α2 + β2 )
xn−1 eαx sin(βx) or xn−1 eαx cos(βx) [D2 − 2αD + (α2 + β2 )]n
I Factors of a linear differential operator with constant coefficients commute.
I The differential opertor that annihilates a function is NOT unique. For example,
(D + 3), (D + 3)(D + 1) and D3 (D + 3) annihilate the function e−3x .
I If T 1 annihilates the function f1 and T 2 annihilates the function f2 , then the product
T 1 T 2 annihilates the sum c1 f1 (x) + c2 f2 (x). For example, D2 annihilates (7 − x)
and D2 + 16 annihilates sin 4x. Then, D2 (D2 + 16) annihilates the linear
combination (7 − x) + 6 sin 4x.
Example: Annihilator Method

Find the particular solution y p of y00 + y = e2x + 1. Also write its general solution.
Step 1: Finding the Complimentary Function yc
Consider the associated homogeneous equation

y00 + y = 0.

Then, solving the auxiliary equation r2 + 1 = 0, we get r1 = i and r2 = −i.

Therefore, the complementary function yc is given by

yc (x) = c1 cos x + c2 sin x,

where c1 and c2 are arbitrary real constants.

Continuation of Previous Slide

Step 2: Finding the Annihilator operator M and operate on L(y) = F .

Observe that (D − 2)e2x = 0 and D(1) = 0. Therefore, D(D − 2)(e2x + 1) = 0 and hence
an annihilator operator is
M = D(D − 2).
Now, M (L(y)) = M (F) gives that

D(D − 2)(D2 + 1) = 0.

Step 3: Finding the general solution of the new homogeneous equation

The general solution of the homogeneous equation D(D − 2)(D2 + 1) = 0 is given by

Yc (x) = c1 cos x + c2 sin x + c3 e2x + c4 ,

where c1 , c2 , c3 and c4 are arbitrary real constants.

Continuation of Previous Slide
Step 4: Forming y p (x) by deleting from Yc (x) all those terms that duplicated in yc (x).
Observe that
Yc (x) = c1 cos x + c2 sin x + c3 e2x + c4 =⇒ Ker(ML) = {cos x, sin x, e2x , 1}.
yc (x) = c1 cos x + c2 sin x =⇒ Ker(L) = {cos x, sin x}.
=⇒ y p (x) = Ae2x + B,
where A and B are real constants to be determined.
Step 5: Substitute y p (x) in L(y) = F and determine the unknown coeficients

y00p + y p = e2x + 1 =⇒ 4Ae2x + Ae2x + B = e2x + 1.

1 1 2x
=⇒ A= and B=1 =⇒ y p (x) = e + 1.
5 5
Step 6: Writing the general solution of L(y) = F
1
General Solution y(x) = c1 cos x + c2 sin x + e2x + 1.
5
Students to Workout in the Class

Using the annihilator method, Find a particular solution of y00 + 4y = cos x.

Answer:
1
y p (x) = cos x .
3
 Operator Method
for
Nonhomogeneous ODEs with Constant
Coefficients
Nonhomogeneous term is a simple function for integrating it.

Refer:
Page 94 to 101 of An Introduction to Differential Equations by Earl A. Coddington, PHI
Learning Pvt. Ltd. New Delhi, 2011.
Operator Method
This method provides a procedure for finding a particular solution y p of
nonhomogeneous linear differential equation L(y) = F with constant coefficients,
where the nonhomogeneous term F is of specific form (Polynomial, Exponential
function, Sine and Cosine functions).
Consider the n-th order nonhomogeneous linear ODE with constant coefficients
dn y dn−1 y dy
L(y) := n + an−1 n−1 + · · · + a1 + a0 y = F(x),
dx dx dx
where a0 , . . ., an−1 are real constants. Note that one can make the coefficient of the
highest order derivative term of the ODE with constant coefficients as 1.
dk
By denoting Dk = dx k , let

P(D) = Dn + an−1 Dn−1 + · · · + a1 D + a0 .

L = P(D).
With P(D), associate a polynomial P(r) = rn + an−1 rn−1 + · · · + a1 r + a0 called the
auxiliary polynomial of P(D).
Note: Throughout, P(D) = Dn + an−1 Dn−1 + · · · + a1 D + a0 actually means that
P(D) = Dn + an−1 Dn−1 + · · · + a1 D + a0 I , where I is the identity operator. In the linear
differential operator expressions, the scalar term a0 actually means a0 I .
Continuation of Previous Slide
Factorize the auxiliary polynomial P(r) as a product of n linear factors as follows:

P(r) = (r − r1 )(r − r2 ) · · · (r − rn ).

Then, P(D) can be factorized as

P(D) = (D−r1 )(D−r2 ) · · · (D−rn ) meaning P(D) = (D−r1 I)(D−r2 I) · · · (D−rn I).

The given nonhomogeneous equation can be written as (P(D))(y(x)) = F(x).

1
The idea behind this method is to define an inverse operator to P(D), written as P(D) (or
(P(D))−1 or P−1 (D)) so that the function y p (x) in the following relations have a meaning.

1
y p (x) = F(x).
P(D)
1
P(D) F(x) = F(x).
P(D)
Inverse Operators of D and (D − r)

Observe that Z
Dy p (x) = F(x) =⇒ y p (x) = F(x) dx.

It is natural to define Z
1
F(x) := F(x) dx.
D

Observe that
1
(D − r) y p (x) = F(x), (r is a constant) =⇒ y p (x) = F(x).
D−r
The solution R of the fist order linear ODE (D − r) y p (x) = F(x) is
y p (x) = erx e−rx F(x) dx. Therefore, we define
Z
1
F(x) := erx e−rx F(x) dx.
D−r
Technique 1: Successive Integrations

P(D) y p = F(x),
where P(D) = (D − r1 )(D − r2 ) · · · (D − rn ).
Then
1 1
y p (x) = F(x) = F(x)
P(D) (D − r1 )(D − r2 ) · · · (D − rn )
!!!
1 1 1
= ··· F(x) .
D − r1 D − r2 D − rn
Apply the operators
Z
1
g(x) = erk x e−rk x g(x) dx for k = 1, 2, . . . , n
D − rk
successively and compute y p (x).
Note: The successive integrations are likely to become complicated and
time-consuming.
Example for Successive Integrations Technique
Using the operator method, find a particular solution of y00 − 3y0 + 2y = xe x .

P(D)y = (D2 − 3D + 2)y = (D − 2)(D − 1)y = xe x .

The particular solution y p is given by
!
1 1 1
y p (x) = xe x = xe x
(D − 2)(D − 1) D−2 D−1
" Z # " Z #
1 1
= e x e−x xe x dx = ex x dx
D−2 D−2
x2 e x
!
1
=
D−2 2
2x
" Z # "Z #
1 2x e
= e e−2x x2 e x dx = x2 e−x dx
2 2
e2x h −x 2 i (−1)
= −e (x + 2x + 1) = (1 + x)2 e x
2 2
Technique 2: Partial Fractions and Integrations

If the factors of P(D) = (D − r1 ) · · · (D − rn ) are distinct, then we can decompose the

1
operator P(D) into partial fractions as

1 A1 A2 An
= + + ··· + .
P(D) (D − r1 ) (D − r2 ) (D − rn )
!
1 A1 A2 An
y p (x) = F(x) = + + ··· + F(x)
P(D) (D − r1 ) (D − r2 ) (D − rn )
1 1 1
= A1 F(x) + A2 F(x) + · · · + An F(x)
(D − r1 ) (D − r2 ) (D − rn )
Z ! Z ! Z !
= A1 er1 x e−r1 x F(x) dx + A2 er2 x e−r2 x F(x) dx + · · · + An ern x e−rn x F(x) dx
Example for Partial Fractions and Integrations Technique
Using the operator method, find a particular solution of y00 − 3y0 + 2y = xe x .

P(D) = (D2 − 3D + 2) = (D − 2)(D − 1).

1 1 1 1
= = − .
P(D) (D − 2)(D − 1) (D − 2) (D − 1)

1 1 1
xe x = xe x − xe x
P(D) (D − 2) (D − 1)
Z Z
= e2x e−2x xe x dx − e x e−x xe x dx
Z Z
= e2x xe−x dx − e x x dx
2 x 2
" #
x e x
= e2x −xe−x − e−x − = (−1) e x 1 + x +
 
2 2
x2
h i
y p (x) = (−1) e 1 + x + 2 is a particular solution of given nonhomogeneous linear
x

ODE.
Operator Method when F(x) = xk
To find a particular solution of P(D)y = xk , where P(D) = Dn + an−1 Dn−1 + · · · + a1 D + a0
with a0 , 0 and k is a positive integer, it is preferable to expand the operator
P(D) = (a0 + a1 D + · · · + an−1 D
1 n−1 + Dn )−1 into power series expansion and operate on

the function xk .
In the power series expansion, it is obvious that terms of the expansion beyond k-th
power of D need not be written, since Dm xk = 0 for all m > k.
To expand into power series, use the Geometric Series.
1 1
= 1 + D + D2 + D3 + · · · and = 1 − D + D2 − D3 + · · ·
1−D 1+D
! " #!−1
1 1 a1 an−1 n−1 1 n
= 1+ D + ··· + D + D
P(D) a0 a0 a0 a0

!  " # " #2 
1  a1 an−1 n−1 1 n a1 an−1 n−1 1 n 
= + + + + + + + + .

1 − D · · · D D D · · · D D · · ·
 
a0 a0 a0 a0 a0 a0 a0


 


If a0 = 0 and a1 , 0, then
! " #!−1
1 1 a2 a3 an−1 n−2 1 n−1
= 1+ D + D2 + · · · + D + D .
P(D) a1 D a1 a1 a1 a1
Example for Operator Method when F(x) = xk

Using the operator method, find a particular solution of y000 − 2y00 + y = x4 + 2x + 5.

P(D) = 1 − 2D2 + D3 .
1 1 1
= = = 1+(2D2
−D3
)+(2D 2
−D3 2
) +(2D2
−D ) +· · ·
3 3
P(D) 1 − 2D2 + D3 1 − (2D2 − D3 )

y p (x) = (1 + 2D2 − D3 + 4D4 + · · · )(x4 + 2x + 5)

Note that the terms Dk for k ≥ 5 are NOT required because Dk (x4 + 2x + 5) = 0 for
k ≥ 5.
= (x4 + 2x + 5) + 2(12x2 ) − (24x) + 4(24)
= x4 + 24x2 − 22x + 101
y p (x) = x4 + 24x2 − 22x + 101 is a particular solution of given nonhomogeneous linear
ODE.
Operator Method when F(x) = Aeαx if P(α) , 0
If P(D) = (D − r) and F(x) = Aeαx , where α and A , 0 are real constants, then
(D − r)Aeαx = A(α − r)eαx .
1
Operating (D−r)(α−r) on both sides of above identity, we get
1 1
Aeαx = Aeαx if r , α.
(D − r) (α − r)
This can be generalized as follows:

Let
P(D) = (D − r1 )(D − r2 ) · · · (D − rn ) and F(x) = Aeαx ,
where α and A , 0 are real constants.
1 αx Aeαx
Ae = if P(α) , 0.
P(D) P(α)
Operator Method when F(x) = Aeαx if P(α) = 0
If P(D) = (D − α)k and F(x) = Aeαx , where α and A , 0 are real constants, and k ∈ N,
then
1 αx xk αx
Ae = Ae .
(D − α)k k!
Let P(D) = Dn + an−1 Dn−1 + · · · + a1 D + a0 .
Suppose that α is a root of the auxiliary polynomial
P(r) = rn + an−1 rn−1 + · · · + a1 r + a0 = 0 with multiplicity k. Then P(r) = (r − α)k Q(r),
where Q(α) , 0. Note that P(k) (α) = k! Q(α).

Let
P(D) = (D − α)k Q(D) and F(x) = Aeαx ,
where α and A , 0 are real constants and Q(α) , 0. Note that P(α) = 0. Then

1 αx 1 αx xk Aeαx xk A eαx
Ae = Ae = = .
P(D) (D − α) Q(D)
k k! Q(α) (k)
P (α)
Example for Operator Method when F(x) = Aeαx if P(α) , 0

Using the operator method, find a particular solution of y000 − y00 + y0 + y = 3e−2x .

P(D) = 1 + D − D2 + D3 .
Note that P(−2) = 1 + (−2) − (−2)2 + (−2)3 = −13 , 0.
We know that αx
1 Ae
Aeαx = if P(α) , 0.
P(D) P(α)
Therefore,
1 e−2x e−2x (−3) −2x
3e−2x
= 3 = 3 = e .
P(D) P(−2) (−13) 13
(−3) −2x
y p (x) = 13 e is a particular solution of given nonhomogeneous linear ODE.
Example for Operator Method when F(x) = Aeαx if P(α) = 0
Using the operator method, find a particular solution of (D − 1)(D + 5)(D − 2)3 y = 3e2x .

P(D) = (D − 2)3 Q(D), where Q(D) = (D − 1)(D + 5).

Note that P(2) = 0 and Q(2) = (2 − 1)(2 + 5) = 7 , 0.
We know that, if

P(D) = (D − α)k Q(D) and F(x) = Aeαx ,

where α and A , 0 are real constants and Q(α) , 0, then

1 αx xk Aeαx
Ae = .
P(D) k! Q(α)
Therefore
1 x3 3e2x x3 3e2x 1 3 2x
3e =
2x
= = x e .
P(D) 3! Q(2) 6 7 14
y p (x) = 1
14 x3 e2x is a particular solution of given nonhomogeneous linear ODE.
Operator Method when P(D) = φ(D2 ) and F(x) = A sin(αx) or A cos(αx)

(D2 )n A sin(αx) = A (−α2 )n sin(αx) and (D2 )n A cos(αx) = A (−α2 )n cos(αx) for n ∈ N.
Let P(D) = φ(D2 ), where φ(T ) is a polynomial function in D2 .
Let F(x) = A sin(αx) or A cos(αx), where α and A , 0 are real constants such that
P(−α) = φ(−α2 ) , 0. Then,
φ(D2 ) A sin(αx) = A φ(−α2 ) sin(αx)
φ(D2 ) A cos(αx) = A φ(−α2 ) cos(αx)

If φ(−α2 ) , 0, then

1 sin(αx)
A sin(αx) = A
φ(D2 ) φ(−α2 )
1 cos(αx)
A cos(αx) = A
φ(D2 ) φ(−α2 )
Example for Operator Method when P(D) = φ(D2 ) and F(x) = A sin(αx)
or A cos(αx)

Using the operator method, find a particular solution of y00 − 4y = 2 sin(x/2).

Here A = 2, α = 1/2 and

P(D) = D2 − 4 = φ(D2 ), where φ(T ) = T − 4.

Observe that φ(−(1/2)2 ) = φ(−1/4) = − 41 − 4 = −17

4 , 0.
sin(αx)
If φ(−α2 ) , 0, then φ(D
1
2) A sin(αx) = A φ(−α2 )
.
Therefore,
   
x
1   x  1   x  sin 2 sin 2x (−8)  x
2 sin = 2 sin =2 =2 = sin
P(D) 2 φ(D2 ) 2 φ(−(1/2)2 ) (−17/4) 15 2
(−8)
y p (x) = 15 sin (x/2) is a particular solution of given nonhomogeneous linear ODE.
Important Note: If P(D) = (D − r1 )(D − r2 ) · · · (D − rn ) is a polynomial of Differential
Operator, where r1 , . . ., rn are real constants/ roots, then P(D) can be written as

φ(D2 )
P(D) = ,
ψ(D)
where φ and ψ are some polynomials with real coefficients.
It gives that
1 ψ(D)
= .
P(D) φ(D2 )

If φ(D2 ) is a polynomial differential operator with φ(−α2 ) , 0, then

1 sin(αx + β)
A sin(αx + β) = A ,
φ(D )
2 φ(−α )
2

1 cos(αx + β)
A cos(αx + β) = A .
φ(D2 ) φ(−α2 )
 γx 
(γx)+i(αx+β)
=e cos(αx + β) + i sin(αx + β)

Using the fact that e
I If the nonhomogeneous term is F(x) = eγx cos(αx + β), then we can find its
particular solution as the real part of particular solution of the ODE having
nonhomogeneous term G(x) = e(γx) + i (αx+β) .
I If the nonhomogeneous term is F(x) = eγx sin(αx + β), then we can find its
particular solution as the imaginary part of particular solution of the ODE having
nonhomogeneous term G(x) = e(γx) + i (αx+β) .

If P(D) is a polynomial differential operator with real coefficients, then

" #
1 1
K eγx cos(αx + β) = K × Real Part of e(γx) + i (αx+β) ,
P(D) P(D)
" #
1 1
K eγx sin(αx + β) = K × Imaginary Part of e(γx) + i (αx+β) .
P(D) P(D)

Here, α, β, γ and K are real constants.

Students to Workout in the Class

Using the operator method, find a particular solution of y00 − y0 + y = e2x cos(x).
Hint:
1 1
= .
P(D) D2 − D + 1

P(2 + i) = (2 + i)2 − (2 + i) + 1 = 2 + 3i , 0.

1 (2+i)x 1 e(2+i)x (2 − 3i) e(2+i)x

e = e (2+i)x
= = .
P(D) P(2 + i) 2 + 3i 13
Answer: y p (x) = 2 2x
13 e cos(x) + 3 2x
13 e sin(x).
Students to Workout in the Class

Using the operator method, find a particular solution of y000 + y00 − y0 − y = cos(2x).
Idea:
1 1 1 (D − 1)
= = .
D + D − D − 1 D + 1 D − 1 (D − 1)
3 2 2 2 2

(−2) cos(2x)
Answer: 25 sin(2x) − 25 .
Other Useful Formulas when F(x) = eαx f (x) or x f (x)

Let P(D) = Dn + an−1 Dn−1 + · · · + a1 D + a0 , where a0 , . . ., an−1 are real constants.

Let f (x) be a sufficient number of times differentiable function.
( )
1 1
(eαx f (x)) = eαx f (x) .
P(D) P(D + α)
0
" # " #!
1 P (D) 1
(x f (x)) = x − f (x) .
P(D) P(D) P(D)
 Method of Variation of Parameters
for
Nonhomogeneous Linear ODEs
This method is more general and also applicable for the linear ODEs with
variable coefficients.
Method of Variation of Parameters
Consider the nonhomogeneous differential equation

dn y dn−1 y dy
n
+ a n−1 (x) n−1
+ · · · + a1 (x) + a0 (x) y = F(x), (4)
dx dx dx
where a0 (x), a1 (x), . . ., an−1 (x) and F(x) are continuous on an interval I = (a, b) ⊆ R.
Note: The coefficient functions are numbered in reverse order and the coefficient of
highest order term is 1.
The general solution of (4) is given by

y(x) = yc (x) + y p (x),

where yc (x) is the complementary function and y p (x) is a particular solution of (4).
The complementary function yc (x) is the general solution of corresponding
homogeneous equation

dn y dn−1 y dy
+ a n−1 (x) + · · · + a1 (x) + a0 (x) y = 0. (5)
dxn dx n−1 dx
Continuation of Previous Slide
Let {y1 , y2 , . . . , yn } be a fundamental set of solutions of the corresponding
homogeneous equation (5). Then,
yc (x) = c1 y1 (x) + c2 y2 (x) + · · · + cn yn (x),
where c1 , . . ., cn are arbitrary real constants.
In the method of Variation of Parameters, we seek a particular solution y p (x) of the form
y p (x) = v1 (x)y1 (x) + v2 (x)y2 (x) + · · · + vn (x)yn (x) (6)
and try to determine v1 (x), v2 (x), . . ., vn (x).
Differentiating y p , we get
n
X n
X
y0p = vi y0i + v0i yi .
i=1 i=1
To avoid second and higher order derivatives of vi ’s, we impose the condition
n
X
v0i yi = 0. (7)
i=1
Continuation of Previous Slide
Therefore,
n
X
y0p = vi y0i .
i=1
Again, differentiating y0p , we obatin
n
X n
X
y00p = i +
vi y00 v0i y0i .
i=1 i=1
For the same reason, we impose the condition
n
X
v0i y0i = 0. (8)
i=1
(n−1)
p , . . ., y p
Likewise, on computing y000 , we impose (n − 3) additional conditions involving
v0i ’s as
n
X n
X
i = 0,
v0i y00 ··· , v0i y(n−2)
i = 0. (9)
i=1 i=1
Continuation of Previous Slide

Finally, for the nth condition, use the derived conditions and the fact that L(yc ) = 0 and
L(y p ) = F to obtain
n
X
v0i y(n−1)
i = F. (10)
i=1

Therefore, we seek v01 , . . . , v0n that satisfy the system

y1 v01 + · · · + yn v0n = 0,
y01 v01 + · · · + y0n v0n = 0,
.. .. .. ..
. . . = .
y(n−1)
1 v 0
1 + · · · + y (n−1) 0
n vn = F.
Continuation of Previous Slide

The existence of v01 , . . . , v0n requires

y1 y2 ··· yn
y01 y02 ··· y0n
.. .. .. = W(y1 , . . . , yn )(x) , 0 on (a, b), (11)
. . ··· .
y(n−1)
1 y (n−1)
2 ··· yn(n−1)

which is true as {y1 , . . . , yn } is a fundamental set of solutions.

On solving for v01 , . . . , v0n , we find that

F(x) Wk (x)
v0k (x) = , for k = 1, 2, . . . , n, (12)
W(y1 , . . . , yn )(x)

where the determinant Wk (x) is obtained from W(y1 , . . . , yn )(x) by replacing k-th
column by [0, . . . , 0, 1]T .
Continuation of Previous Slide

We can express Wk (x) as

Wk (x) = (−1)(n−k) W(y1 , . . . , yk−1 , yk+1 , . . . , yn )(x) for k = 1, . . . , n.

Integrating v0k (x), we get

Z
F(x) Wk (x)
vk (x) = dx, for k = 1, 2, . . . , n. (13)
W(y1 , . . . , yn )(x)
Then, by (6),
n n Z !
X X F(x)Wk (x)
y p (x) = vk (x)yk (x) = yk (x) dx .
k=1 k=1
W(y1 , . . . , yn )(x)
For n = 2, Method of Variation of Parameters

y00 + a1 (x)y0 + a0 (x) y = F(x).

Important Note: The Coefficient of Highest Order Derivative is 1.

yc (x) = c1 y1 (x) + c2 y2 (x).

y1 (x) y2 (x)
W(y1 , y2 )(x) = 0 0 = y1 (x)y02 (x) − y2 (x)y01 (x).
y1 (x) y2 (x)
0 y2 (x) y1 (x) 0
W1 (x) = = −y2 (x), W2 (x) = = y1 (x).
1 y02 (x) y01 (x) 1
By (6) and (13),
y p (x) = v1 (x)y1 (x) + v2 (x)y2 (x), where (14)
Z Z
(−1)F(x) y2 (x) F(x) y1 (x)
v1 (x) = dx and v2 (x) = dx. (15)
W(y1 , y2 )(x) W(y1 , y2 )(x)
Example for Method of Variation of Parameters
Find a particular solution of y00 + y = cosec(x). Also write its general solution.
Observe that yc (x) = c1 y1 (x) + c2 y2 (x), where y1 (x) = sin x and y2 (x) = cos x; and c1 , c2
are arbitrary real constants.

y1 (x) y2 (x) sin x cos x

W(y1 , y2 )(x) = 0 0 = = − sin2 x − cos2 x = −1 , 0
y1 (x) y2 (x) cos x − sin x
0 cos x sin x 0
W1 (x) = = − cos x, W2 (x) == sin x.
1 − sin x cos x 1
Z Z
(−1)cosec(x) cos(x) cosec(x) sin(x)
v1 (x) = dx = log(sin x) and v2 (x) = dx = −x.
(−1) (−1)

y p (x) = v1 (x)y1 (x) + v2 (x)y2 (x) = log(sin x) sin x − x cos x.

y(x) = yc (x) + y p (x) = c1 sin x + c2 cos x + log(sin x) sin x − x cos x.

Students to Workout in the Class (Variation of Parameters)
Find the general solution of

x2 y00 − 6x y0 + 10 y = 3x4 + 6x3 ,

given that y1 (x) = x2 and y2 (x) = x5 are linearly independent solutions of the
corresponding homogeneous equation.
Answer: Note that x2 y00 − 6x y0 + 10 y = 0 is in the form of Cauchy-Euler equation and
we will study next how to solve the Cauchy-Euler equation. Right now, it is already
given that y1 (x) = x2 and y2 (x) = x5 are its linearly independent solutions.

3 4
y p (x) = −3x − x .
3
2
3 4
y(x) = c1 x + c2 x − 3x − x ,
2 5 3
2
where c1 and c2 are arbitrary real constants.