1.

13 Higher Order Linear Differential Equations

The general linear differential equation of order n is an equation that can be written as
……………….(1)
If R(x) = 0, then the equation is called a homogenous linear differential equation otherwise it is
called non-homogeneous differential equation. The coefficient functions are
continuous on the interval I.
Consider .
If and are solutions of homogenous equation then is also a solution of
that equation.
In a similar manner, if y1 ,y2 ,…,yn are solutions of (1), then
……..(2)
is also a solution where are arbitrary constants.
The expression (2) is called linear combination of the functions .
For a homogeneous equation, the
general solution is of the form where are all possible
linearly independent solutions of the given equation. Where as for the nonhomogeneous equation
the general solution is of the
form y = yc + y p , where yc is the general solution of the corresponding homogenous

equation and y p is a particular

solution of the given equation , which does not contain any arbitrary constants.
1.3.1 An Existence and Uniqueness Theorem:
Given an order linear differential equation,
on an interval I,
suppose that is any number on the interval I and are n arbitrary constants.
Example 1.13.1 Find the unique solution of , .
Solution is .
Using we get .

1.13.2 Linear Independence of Solutions:

Given the functions if constants not all zero, exist such that
for all in , then the functions
are said to be linearly dependent on that interval. If no such relation exists, the functions are said
to be linearly independent.
1.13.3 The Wronskian of Solution:

25
To test whether n functions are linearly independent on an interval let us assume that
each of the functions is differentiable atleast ( times .
Then from the equation c1 f1 + c2 f 2 + ... + cn f n = 0 , it follows by successive differentiation that
c1 f1' + c2 f '2 + ... + cn f n ' = 0
c1 f1'' + c2 f 2+ ... + cn f n = 0
…
c1 f1n−1 + c2 f 2n−1 + ... + cn f n n−1 = 0
For any fixed value of in , the nature of solutions of these will be determined by the
f1 ( x) f 2 ( x) ... f n ( x)
f ' ( x) f 2 ' ( x) ... f n ' ( x)
determinant W ( x) = 1
. . . .
f1n −1 ( x) f 2 n −1 ( x) ... f n n −1 ( x)
If for some on , then c1 = c2 =…=cn = 0 Hence, functions are linearly
independent on . The function is called Wronskian of the functions .
Example 1.13.2 Let y1 = eax and y2 = ebx , then y1 = aeax , y2 = bebx .
eax ebx
Then wronskian is given by W = = ( b − a ) e( a +b ) x = 0 only if a=b.
ax bx
ae be
Therefore for a≠b, y1 = eax and y2 = ebx are linearly independent.
Example 1.13.3. Let y1 = 1 and y2 = x , then y1 = 0, y2 =1
1 x
Then wronskian is given by W = =1 ≠ 0. Therefore y1 = 1 and y2 = x are linearly
0 1
independent.
d dk
Let D = then D = k for k =1, 2,.... .Then the equation (1) can be expressed as
k
dx dx
b0 Dn y + b1D n−1 y + ... + bn y = R( x)
i.e., (b0 D n + b1D n−1 + ... + bn ) y = R( x)
i.e., f(D) y = R(i) where f(D) = b0 D n + b1D n−1 + ... + bn is called a differential operator. To
solve such equations we first study the properties of the differential operator.
1.13.4 Properties of differential operator:
1. f(D)eax = eaxf(a)
Proof: Let f(D) = b0 D n + b1D n−1 + ... + bn .
Since Dkeax = akeax, for k = 1,2,3…n. we have,

26
 f(D)eax = b0 D n eax + b1D n −1e ax + ... + bn e ax
= b0 a n e ax + b1a n−1e ax + ... + bn e ax
= (b0 a n + b1a n −1 + ... + bn )e ax = f (a )e ax
2. f(D)eax y = eaxf(D+a)y
Proof: Let f(D) = b0 D n + b1D n−1 + ... + bn

We have D eax y = y Deax + eax Dy = y eax a + eax Dy = eax (D+a)y,

 D2(eax y) = D(D(eax y)) = D(eax (D+A)y) = eax (D+a)(D+a)y= eax (D+a)2y
Similarly we can show that Dk(eax y) = eax (D+a)ky, k = 1,2,...n.
Substituting in the formula we get the required result.

0, j = 0,1, 2,...k − 1
3. ( D − a)k eax x j =  ax
e k !, j = k


0, j = 0,1, 2,....k − 1

Proof: We know that D k x j =  then by property(2), the result follows.
k !, j = k
1.13.5 The solution of linear homogeneous differential equation with constant coefficients

Consider the linear homogeneous differential equation with constant coefficients

dny d n−1 y dy
b0 n + b1 n−1 + ... + bn−1 + bn y = 0 where are all constants.
dx dx dx
This equation can be rewritten as f(D)y = 0 where f(D) = b0 D n + b1D n−1 + ... + bn .

If y = eax , then by property (1), f(D)y = f(D)eax = eaxf(a) = 0  f(a) = 0.

This equation is called the auxiliary or characteristic equation associated with the given
differential equation. For a nth order differential equation, the auxiliary equation has n roots say,
a1 , a2 ,......an . Then y1 = ea1x , y2 = ea2 x ,..........., yn = e an x are all solutions of the given differential

equation.
Case1: If the roots of the auxiliary equations are all distinct then
y1 = ea1x , y2 = ea2 x ,..........., yn = e an x are linearly independent solutions of the given equation.

Hence the general solution is given by y = C1ea1x + C2ea2 x + ... + Cne an x .

Case2: If some roots are equal, say a1 = a2 = ...... = ak = a.

27
Then the solutions y1 = y2 = ... = yk = e ax . Then the solution is given by y = ea1x + ea2 x + ... + ean x

does not contain n arbitrary constants and hence cannot be the general solution. Since first k
roots of the auxiliary equations are equal the given differential equation can be rewritten as
g ( D)( D − a)k y = 0. Then by property (3), we observe that y j = eax x j , j = 0,1,......,.k − 1 are all

solutions of the given equation. Hence the general solution is

y = (C1 + C2 x + C3 x 2 + ...... + Ck x k −1 )e ax + Ck +1e ak +1x + ...... + Cne an x .

Case3. If some roots of the auxiliary equation are real. Since for equation with real coefficients
the roots exists in conjugate pairs, let 1 = a + ib and  2 = a − ib be two roots. Then

y1 = e1x = e( a +ib) x = eax (cos bx + i sin bx) and y2 = e 2 x = e( a −ib ) x = e ax (cos bx − i sin bx) are the two

distinct solutions. Therefore the corresponding solution is y = C1e1x + C2e 2 x , which can be

expressed as y1 = eax ( A1 cos bx + A2 sin bx) where A1 and A2 are arbitrary constants.

Example 1.13.4: Solve (2D2 + 5D – 12)y = 0

Solution: Auxiliary equation is 2m2 + 5m – 12 = 0.
That is, (2m – 3) (m + 4) = 0 and it has the roots m1 = 3/2 and m2 = - 4 which are real and
distinct. Therefore the general solution is y = C1 e3x/2 + C2e-4x
d2y
Example 1.13.5 Solve +4y=0
dx 2
Solution: Auxiliary equation is m2 + 4 = 0 or m2= - 4
Therefore m =  2i = 0  2i
Therefore y = C1 cos 2x + C2 sin 2x
d 3x d2y dy
Example 1.13.6 Solve 3
+ 4 2
+4 = 0.
dx dx dx
Solution: Auxiliary equation is m3 + 4m2 + 4m = 0.
That is, m(m + 2)2 = 0
Therefore m1 = 0, m2 = -2, m3 = -2 are its roots.
Here we find that two roots are equal. Hence the general solution is given by
y = C1 e0x + (C2 + C3x) e-2x
Or y = C1 + (C2 + C3i) e-2x

28
 d 4x d2y
Example 1.13.7 Solve 4 + 8 + 16y = 0
dx dx 2
Solution: Here the auxiliary equation is m4 + 8m2 + 16 = 0, which is simply (m2 + 4)2 = 0.
Therefore we have m2 + 4 = 0 repeated. It gives m2 = -4
Therefore m =  2i.
Thus the roots are m =  2 i,  2i (imaginary, repeated).
Hence the general solution is y = e0x [(C1 + C2x) cos 2x + (C3 + C4x) sin 2x]
That is, y = (C1 + C2x) cos 2x + (C3 + C4x) sin 2x.
d4y d3y d2y dy
Example 1.13.8 Solve 4
-2 3
+2 2
-2 +y=0
dx dx dx dx
Solution: Hence the A.E. is m4 – 2m3 + 2m2 – 2m + 1 = 0.
The roots are m = 1, 1,  i
Therefore the general solution is y = (C1 + C2x) ex + e0x (C3 cos x + C4 sin x)
That is, y = (C1 + C2x)ex + (C3 cos x + C4 sin x).

1.13.6 The solution of linear non-homogeneous differential equation with constant

coefficients

We know that the general solution of nonhomogeneous linear differential equation

is of the form y = yc + y p , where yc is the general solution of the corresponding homogenous

equation , called the complementary function and

y p is a particular solution of the given equation , which does not contain any arbitrary constants.

The complementary function can be determined using the method described above. To determine
the particular solution y p , we use the following methods.
1.13.6.1 Inverse differential operator method
1
If f(D)y =  (x) then we define the inverse differential operator denoted by as
f ( D)
1 d
[ (x)] = y , where ‘D’ is the differential operator .
f ( D) dx

29
 1
Thus f(D) is also a differential operator and can be treated as its inverse.
f ( D)

1
Example 1.13.6.1 y =  ydx DR( x) = y  R( x) = ydx.
D
1 1 11 1 x2
 1 =  1.dx = x , 2 1 =  1 = x =  xdx =
D D DD D 2
1 xk
Similarly we can showthat 1 = , k = 1, 2,3,.......
Dk k!
1.13.6.2. Properties of the inverse differential operator:
1 eax
2. e ax = if f (a )  0.
f ( D) f (a)
Proof: We know that f(D)eax = eaxf(a), If f(a)  0, then dividing by f(a) we get
1 eax
f ( D)eax = f (a)
f (a) f (a)
1 eax
 e =
ax
f ( D) f (a)
1 1
3. eax y = eax y
f ( D − a) f ( D)
Proof: From the property f(D)eax y = eaxf(D+a)y the result follows.
1
4. y = eax  e−ax y dx
D−a
Proof: The result follows directly from the result (2) and example (1) .
To determine the particular solution of a linear non-homogenous differential equation, we use
1
inverse differential operators. If f(D)y =  (x) ,then yp = [ (x)]
f ( D)

eax
Case(i): If  (x) = eax , then yp = if f(a)  0.
f (a)

If f(a) = 0, then f(D) can expressed as f(D) = (D-a)k  (D), where  (a)  0.for k = 1,2,3,….

1 ax 1 1  1 ax  1  eax 
Then e = eax = k 
e =  
f(D) (D-a)  (D)
k
(D-a)  (D)  (D-a)k  (a) 
1  1 ax  eax 1 eax x k
=  e  = 1 =
 (a)  (D-a)k   (a) D
k
 (a) k !

30
Case(ii): If (x)= cosax or sinax then, since cosax = Re( eiax ) and sinax = Im( eiax ) this case
reduces to the case(i) and hence can be solved.
1
Case(iii) If (x)= xm, for some positive integer m, then can be expanded as a series in
f(D)
1 m
positive powers of x and hence x can be determined.
f(D)
Working Rule:
1. If (x) = eax, then
1 eax
eax = if f (a )  0.
f ( D) f (a)
1 1 eax
If f (a) = 0 then e ax = x e ax = x if f (a )  0,
f ( D) f ( D) f (a)
1
= x2 eax if f (a ) = 0.
f ( D)
1
= x2 eax if f (a ) = 0 and so on...
f (a )
2. If (x) = cos ax or sin ax and if the differential operator can be written as f(D2) then
1 1
[(x)] = (x) provided f(-a2)  0.
2
f (D ) f (−a 2 )
1 1
If f(-a2) = 0 and f (-a 2 )  0, then ( (x)) = x. ( (x))
2
f(D ) f  (-a 2 )

d2y dy
Example 1.13.8 Solve 2
-6 + 10 y = cos 2x + e-3x
dx dx
Solution: Auxiliary equation is m2 – 6m + 10 = 0.

6  36 − 40 6  −4 6  2i
Therefore m = = = =3 i
2.1 2 2
Or m =  i, where  = 3 and  = 1.
Therefore C.F. = e3x (C1 cos x + C2 sin x)
1 1
P.I. - (cos 2x) + 2 (e-3x)
D − 6 D + 10
2
D − 6 D + 10
Using P.I rule 2 and rule 1 respectively we get,

31
 1 1
P.I. = (cos 2x) + e-3x
− 2 − 6 D + 10
2
(−3) − 6(−3) + 10
2

1 1
= (cos 2x) + .e-3x
6 − 6D 9 + 18 + 10
1 1+ D 1 -3x
= (cos 2x) + e , multiplying numerator and denominator by 1 + D in the first
6 1− D 2
37
expression.
1 1+ D 1 -3x
That is, P.I. = (cos 2x) + e
6 1 − (−2 )
2
37
1 1 -3x
{1 . cos 2x + D(cos 2x)} + e .
30 37
1 1 -3x d
That is, P.I = (cos 2x – 2 sin 2x) + e , since D =
30 37 dx
General solution is y = C.F. + P.I.
1 1 -3x
Therefore y = e3x (C1 cos x + C2 sin x) + (cos 2x – 2 sin 2x) + e
30 37
Example 1.13.9 Solve (D4 + 18 D2 + 81) y = cos3 x.
Solution: Auxiliary equation is m4 + 18 m2 + 81 = 0
That is (m2 + 9)2 = 0. Therefore m = 3i,  3i.
Thus C.F. = e0.x {(C1 + C2 x) cos 3x + (C3 + C4x) sin 3x}
C.F. = (C1 + C2x) cos 3x + (C3 + C4x) sin 3x.
1
P.I = (cos3 x)
D + 18 D + 81
4 2

1 3 1 
=  cos x + cos 3x  .
D + 18 D + 81  4
4 2
4 
3 1
( Since Cos3 A = cos A + cos 3A).
4 4
3 1 1 1
P.I = (cos x) + (cos 3x)
4 D + 18 D + 81
4 2
4 D + 18D2 + 81
4

3 1 1 1
= (cos x) + (cos 3x)
4 (−1 ) + 18(−1) + 81
2 2 2
4 D + 18 D 2 + 81
4

But f(D2) = D4 + 18D2 + 81 and a = 3 in the second expression.

32
We observe that f(-a2) = f(-9) = 81 – 162 + 81 = 0
Also f 1(D) = 4D3 + 36 D + 0 = 4 D(D2 + 9)
Therefore f 1(-9) = 4 D(-9 + 9) = 0, also
1 1
Hence (cos 3x) = x2 11 2 (cos 3x)
D + 18 D + 81
4 2
f (D )
But f 1(D) = 4D3 + 36 D, Therefore f 11(D) = 12 D2 + 36
1 1
Hence (cos 3x) = x2, 11 cos 3x
D + 18 D + 81
4 2
f (−3 2 )

1 1 2
That is, x2 cos 3x = - x cos 3x
12(−9) + 36 72
3 1 1 1 2 
Thus P.I. = . cos x +  − x cos 3x 
4 1 − 18 + 81 4  72 
3 1 2
Therefore P.I. = .cos x - .x cos 3x.
256 288
3 1 2
That is, y = (C1 + C2x) cos 3x + (C3 + C4x) sin 3x + cos x - x cos 3x.
256 288
Example 1.13.10 Solve (D2 – 6D + 9)y = x2 + x + 1
Solution: Auxiliary equation m2 – 6m + 9 = 0.
That is (m – 3)2 = 0. Therefore m = 3, 3
Thus C.F. = (C1 + C2 x)e3x.
1
P.I = (x2 + x + 1)
D − 6D + 9
2

1
That is, = (x2 + x + 1)
 2 1 
91 − D + D 2 
 3 9 

−1
1  2 1 
= . 1 −  D − D 2  (x2 + x + 1)
9  3 9 
(Using the binomial expansion (1 – a)-1 = 1 + a + a2 + …), we have

1 
 2 1 2 2 1 2
2


P.I =  
1 + D − D +
  D − D  + ... (x2 + x + 1)
9 
 3 9  3 9  


33
 1 2 1 4
= {1 + D - D2 + D2 + higher powers} (x2 + x + 1)
9 3 9 9
1 2 1
= (1 + D + D2)(x2 + x+ 1), neglecting D3 and higher powers since we have only 2nd degree
9 3 3
polynomial x2 + x + 1.
1 2 d 2 1 d2 2
Therefore P.I. = {1(x2 + x + 1) + (x + x + 1) + (x + x + 1)}
9 3 dx 3 dx 2
1 2 2 1
= {(x + x + 1) + (2x + 1) + (2)}
9 3 3
1 2 7 7 1
= x + x+ = (3x2 + 7x + 7)
9 27 27 27
1
Thus y = C.F. + P. I. = (C1 + C2x) e3x + (3x2 + 7x + 7).
27
d3y d2y
Example 1.13.11 Solve + 3 = 1 + x + e-3x
dx 3 dx 2
Solution: Auxiliary equation is m3 + 3m2 = 0
Thus C.F. = (C1 + C2x) e0x + C3 e-3x or C1 + C2 x + C3 e-3x
1
P.I. = (1 + x + e-3x)
D + 3D
3 2

1
= (1 + x + e-3x)
 D
3 D 2 1 + 
 3
−1
1  D 1
= 1 +  (1 + x) + x . e-3x
3D 2  3 3D + 6 D
2

1  D D 2 D3  1
= 1 − + − + ...  (1 + x) + x. e-3x
3D 2  3 9 27  3D + 6 D
2

 1 1 1 D 1
= 2− + −  (1 + x) + x. e-3x
 3D 9 D 27 81  3(−3) + 6(−3)
2

1 1 1 1 1 1 x
= . 2 (1 + x) - . (1 + x) + (1 + x) - D(1 + x) + e-2x
3 D 9 D 27 81 9
1 x2 1 x2 x3
But (1 + x) = x + , 2 (1 + x) = + and D(1 + x) = 1.
D 2 D 2 6

34
 1  x2 x3  1  x2  1 1 x
Therefore P.I. =  +  -  x +  + (1 + x) - (1) + e-3x
3  2 6  9  9  27 81 9

2 2 25 2 1 3 x -3x
= - x+ x + x + e
81 27 162 18 9
2 2
The term - x may be neglected in view of the arbitrary C1 + C2x appearing in C.F.
81 27
25 2 1 3 x -3x
General Solution is y = C.F. + P.I. = C1 + C2x + C3 e-3x + x + x + e
162 18 9
d3y d2y dy
Example 1.13.12 Solve 3
+ 2 2
+ = e2x + x3
dx dx dx
Solution: Auxiliary equation is m3 + 2m2 + m = 0
m(m2 + 2m + 1) = 0 or m (m + 1)2 = 0
Therefore m = 0, -1, -1
Thus C.F. = C1 e0x + (C2 + C3 x) e-x or C1 + (C2 + C3 x) e-x
1 2x 1
P.I. = (e ) + (x3)
D + 2D + D
3 2
D + 2D + D
3 2

1 2x 1 2x e 2x
Now, (e ) = .e =
D 3 + 2D 2 + D 2 3 + 2.2 2 + 2 18
1
Consider (x3)
D + 2D + D
3 2

Here f(D) = D3 + 2 D2 + D and f(0) = 0

1 1
Therefore we write, (x3) = (x3)
D + 2D + D
3 2
D( D + 2 D + 1)
2

1 1 1
= . (x3) = (1 + D)-2 (x3)
D (1 + D) 2
D
(Using (1 + a)-2 = 1 – 2 a + 3a2 - + …), we have
1 1
(x3) = (1 – 2D + 3D2 – 4D3 + 5 D4) (x3), neglecting higher powers of D.
D + 2D + D
3 2
D
1
=( - 2 + 3D – 4D2 + 5 D3) (x3)
D
1
Now, stands for integration, and D, D2 etc. for successive derivatives.
D

35
 1 x4
Therefore (x3
) = - 2 .x3 + 3(3x2) – 4 (6x) + 5(6).
D 3 + 2D 2 + D 4
x4
= - 2x3 + 9x2 – 24x + 30
4
1 2x x 4
Thus P.I. = e + - 2x3 + 9x2 – 24x + 30
18 4
The constant term 30 may be neglected in view of the arbitrary constant C1 in the C.F.
1 2x x 4
Hence, the general solution y = C1 + (C2 e-x – C3) e-x + e + - 2x3 + 9x2 – 24x.
18 4
Exercise: 1.13.13
d2y dy
1. Solve 4 2
+ 16 - 9y = 4 ex/2 + 3 sin (x/4)
dx dx
2. Solve (D2 + 1)y = ex + x4 + sin x.
Answers :
1 9
1. Auxiliary equation is 4m2 + 16 m – 9 = 0. The roots are m = and - are the roots.
2 2
The C.F. = C1 ex/2 + C2 e-9x/2
1
P.I. = [4ex/2 + 3 sin (x/4)]
4 D + 16 D + 9
2

1 1
=4 (ex/2) + 3. [sin (x/4)].
4 D + 16 D − 9
2
4 D + 16 D − 9
2

In the first expression f(D) = 4 D2 + 16 D – 9.

2
1 1 1
Therefore f   = 4   + 16   - 9 = 1 + 8- 9 = 0.
2 2 2
1 1
Therefore (ex/2) = x. (ex/2).
4 D + 16 D − 9
2
8D + 16
1
In the second expression, replace D2 by - .
42
1 1
Therefore P.I. = 4x (ex/2) + 3. [sin (x/4)]
8D + 16  1 
4 − 2  + 16 D − 9
 4 

36
 1 1
P.I = 4x. .ex/2 + 3. [sin (x/4)]
1 1
8  + 16 − + 16 D − 9
2 4

37
16 D +
1 4
= xex/2 + 3. 2
[sin (x/4)]
5  37 
256 D 2 −  
 4

 37 
316 D + 
1
Thus P.I = xex/2 +  4 
[sin (x/4)]
2
5  1   37 
256 − 2  −  
 4   4 
1 x/2 12
= xe - (64 D + 37) sin (x/4)
5 1625
General solution is
1 x/2 12
y = C.F + P.I. = C1 ex/2 + C2 e-9x/2 + xe - {16 cos (x/4) + 37 sin (x/4)}
5 1625
2. The roots of the auxiliary equation is m =  i.
Therefore C.F. = e0x (C1 cos x + C2 sin x) that is, = C1 cos x + C2 sin x.
1 1 1
P.I = (ex + x4 + sin x) = ex + (x4 – 12x2 +24) + x . (-cos x )
D +1
2
2 2
1 x 4 x
General solution is y = C1 cos x + C2 sin x + e + x –12 x2 + 24 - cos x.
2 2

37