ODEs

𝑑𝑦
+ 𝑦𝑥 2 = 𝑓(𝑥) is a 1st order, 1st degree Ordinary Differential equation(ODE).
𝑑𝑥

𝑑2𝑥 𝑑𝑥 3
+ 5 ( ) − 4𝑥 = 𝑒 𝑡 is a 2nd order, 1st degree ODE.
𝑑𝑡 2 𝑑𝑡

𝑑𝑦
Example 1 = 𝑦, 𝑦(0) = 3, initial-value problem(IVP).
𝑑𝑥

𝑑𝑦 𝑑𝑦
Solution =𝑦⇒ = 𝑑𝑥
𝑑𝑥 𝑦

Integrating gives 𝑙𝑛𝑦 = 𝑥 + 𝑐 ⇒ 𝑦 = 𝑒 𝑥+𝑐 = 𝑒 𝑥 ∙ 𝑒 𝑐 = 𝑘𝑒 𝑥 (i)

where 𝑒 𝑐 = 𝑘

Put 𝑦(0) = 3 in (i) ⇒ 3 = 𝑘𝑒 0 = 𝑘

Now (i) becomes 𝑦 = 𝑘𝑒 𝑥 ⇒ 𝑦 = 3𝑒 𝑥 .

𝑑𝑦 𝑥
Example 2 =− , 𝑦(4) = −3.
𝑑𝑥 𝑦

𝑑𝑦 𝑥
Solution = − ⇒ 𝑦 𝑑𝑦 = −𝑥𝑑𝑥
𝑑𝑥 𝑦

Integrating
𝑦2 𝑥2
=− + 𝑐 ⇒ 𝑥 2 + 𝑦 2 = 2𝑐 = 𝑘 (ii)
2 2

Put 𝑦(4) = −3 in (ii) gives

42 + (−3)2 = 𝑘 ⇒ 𝑘 = 25

Thus (ii) becomes 𝑥 2 + 𝑦 2 = 25 , solution.

1
 Linear 1st ordered ODEs
𝑑𝑦
𝑎(𝑥) + 𝑏(𝑥)𝑦 = 𝑓(𝑥) is a linear 1st order ODE.
𝑑𝑥

𝑑𝑦 𝑏(𝑥) 𝑑𝑦
and + 𝑦 = 𝑓(𝑥)/𝑎(𝑥) that is + 𝑝(𝑥)𝑦 = 𝑔(𝑥) ….. (i)
𝑑𝑥 𝑎(𝑥) 𝑑𝑥
is linear standard form .

Equation (i) has solution as sum of two solutions : 𝑦(𝑥) = 𝑦𝑐 + 𝑦𝑝 .

𝑑𝑦
where 𝑦𝑐 is a solution of + 𝑝(𝑥)𝑦 = 0 and 𝑦𝑝 is a particular solution of the
𝑑𝑥
equation (i).
𝑑𝑦
Example 1 Solve the equation 𝑥 − 4𝑦 = 𝑥 6 𝑒 𝑥 .
𝑑𝑥

In standard form
𝑑𝑦 4𝑦
− = 𝑥 5𝑒 𝑥 (ii)
𝑑𝑥 𝑥

−4
Here 𝑝(𝑥) =
𝑥

4
Now find ∫ 𝑝 𝑑𝑥 = ∫ − 𝑑𝑥 = −4 𝑙𝑛𝑥 = 𝑙𝑛𝑥 −4 .
𝑥
−4
So integrating factor (I.F.)= 𝑒 ∫ 𝑝𝑑𝑥 = 𝑒 𝑙𝑛𝑥 = 𝑥 −4 .

Now multiply both sides of equation (ii) by the I.F. and get
𝑥 −4 𝑑𝑦 4𝑦 𝑑
− . 𝑥 −4 = 𝑥 5 𝑒 𝑥 𝑥 −4 ⇒ (𝑦 𝑥 −4 ) = 𝑥𝑒 𝑥 .
𝑑𝑥 𝑥 𝑑𝑥

Integrating gives,

𝑦 𝑥 −4 = ∫ 𝑥𝑒 𝑥 𝑑𝑥 = 𝑥𝑒 𝑥 − 𝑒 𝑥 + 𝑐

So 𝑦 = 𝑥 5 𝑒 𝑥 − 𝑥 4 𝑒 𝑥 + 𝑐𝑥 4 is the solution .

𝑑𝑥
Example 2 Solve 𝑡 + (1 + 𝑡)𝑥 = 𝑒 −𝑡 𝑠𝑖𝑛2𝑡.
𝑑𝑡

2
Solution Standard form
𝑑𝑥 (1+𝑡) 𝑒 −𝑡 𝑠𝑖𝑛2𝑡
+ 𝑥= ( i)
𝑑𝑡 𝑡 𝑡

1+𝑡 1
Here 𝑝(𝑡) = = +1
𝑡 𝑡

1
∴ ∫ 𝑝(𝑡)𝑑𝑡 = ∫ ( + 1) 𝑑𝑡 = 𝑙𝑛𝑡 + 𝑡
𝑡

∴ 𝐼. 𝐹. = 𝑒 ∫ 𝑝(𝑡)𝑑𝑡 = 𝑒 𝑡+𝑙𝑛𝑡 = 𝑒 𝑡 ∙ 𝑒 𝑙𝑛𝑡 = 𝑡 𝑒 𝑡

Multiplying the bothsides of equation(i) by the I.F.

𝑑 𝑒 −𝑡 𝑠𝑖𝑛2𝑡
⇒ (𝑥 ∙ 𝑡𝑒 𝑡 ) = ∙ 𝑡 𝑒 𝑡 = 𝑠𝑖𝑛2𝑡
𝑑𝑡 𝑡

Integrating gives,
1
𝑥 𝑡𝑒 𝑡 = ∫ 𝑠𝑖𝑛2𝑡 𝑑𝑡 = − 𝑐𝑜𝑠2𝑡 + 𝑐, solution.
2

𝑑𝑥 𝑑𝑦
Examples (a) (𝑡 + 1) + 𝑥 = 𝑙𝑛𝑡 (b) + 𝑦 𝑡𝑎𝑛𝑡 = 𝑐𝑜𝑠 2 𝑡
𝑑𝑡 𝑑𝑡

𝑑𝑦 𝑑𝑦
(c ) 𝑥 + 4𝑦 = 𝑥 3 − 𝑥 (d) 𝑐𝑜𝑠𝑥 + (𝑠𝑖𝑛𝑥)𝑦 = 1
𝑑𝑥 𝑑𝑥

𝑑𝑟
(e ) + 𝑟 𝑠𝑒𝑐𝜃 = 𝑐𝑜𝑠𝜃
𝑑𝜃

Bernoulli’s Equations
𝑑𝑦
+ 𝑝(𝑥)𝑦 = 𝑓(𝑥) 𝑦 𝑛 [𝑛 ≠ 0,1] (i)
𝑑𝑥

Solution technique: Use 𝑢 = 𝑦1−𝑛

𝑑𝑦
Example 1 𝑥 + 𝑦 = 𝑥 2𝑦2
𝑑𝑥

Dividing bothsides by 𝑥, we get

𝑑𝑦 𝑦
+ = 𝑥𝑦 2
𝑑𝑥 𝑥

3
 𝑑𝑦 1
⇒ 𝑦 −2 + 𝑦 −1 = 𝑥 (ii)
𝑑𝑥 𝑥

We put 𝑢 = 𝑦 −1
𝑑𝑢 𝑑𝑦
∴ = (−1) 𝑦 −2
𝑑𝑥 𝑑𝑥
Then (ii) becomes
𝑑𝑢 𝑢 𝑑𝑢 𝑢
− + =𝑥⇒ − = −𝑥 which is linear. (iii)
𝑑𝑥 𝑥 𝑑𝑥 𝑥

1
Here 𝑝(𝑥) = −
𝑥

1
∴ ∫ 𝑝(𝑥)𝑑𝑥 = ∫ − 𝑑𝑥 = −𝑙𝑛𝑥
𝑥
−1 1
Now I.F. = 𝑒 −𝑙𝑛𝑥 = 𝑒 𝑙𝑛𝑥 = 𝑥 −1 =
𝑥

Multiplying bothsides oef equation (iii) by the I.F , we get

𝑑 1 1
( 𝑦) = −𝑥 ∙ = −1
𝑑𝑥 𝑥 𝑥
Integrating gives
1
𝑦 = −𝑥 + 𝑐 ⇒ 𝑦 = 𝑐𝑥 − 𝑥 2 , solution .
𝑥

𝑑𝑦 𝑑𝑦
Examples (a) 𝑥𝑦 2 = 𝑦 3 − 𝑥 3 , 𝑦(1) = 2 (b) 𝑥 − (1 + 𝑥)𝑦 = 𝑥𝑦 2
𝑑𝑥 𝑑𝑥

𝑑𝑦
(c ) √𝑦 + 3√𝑦 = 1 , 𝑦(0) = 4
𝑑𝑥

EXACT DIFFERENTIAL EQUATIONS

4
If 𝑧 = 𝑓(𝑥, 𝑦) is a function of variables , 𝑦 , then its total differential :
𝜕𝑓 𝜕𝑓
𝑑𝑧 = 𝑑𝑥 + 𝑑𝑦 (i)
𝜕𝑥 𝜕𝑦

Now if 𝑓(𝑥, 𝑦) = 𝑐 , then (i) becomes

𝜕𝑓 𝜕𝑓
𝑑𝑥 + 𝑑𝑦 = 0 (ii)
𝜕𝑥 𝜕𝑦

For example, if 𝑥 2 − 5𝑥𝑦 + 𝑦 3 = 𝑐, then using (ii) we get

(2𝑥 − 5𝑦)𝑑𝑥 + (−5𝑥 + 3𝑦 2 )𝑑𝑦 = 0.

Now equation (ii) is an exact differential equation if

𝜕 𝜕𝑓 𝜕 𝜕𝑓
( ) = 𝜕𝑥 (𝜕𝑦)
𝜕𝑦 𝜕𝑥
(iii)

𝜕𝑓 𝜕𝑓
So if we let 𝑀 = and 𝑁 = , then
𝜕𝑥 𝜕𝑦

𝜕 𝜕
(ii) becomes 𝑀𝑑𝑥 + 𝑁𝑑𝑦 = 0 and is exact if 𝑀= 𝑁 by (iii).
𝜕𝑦 𝜕𝑥

How to solve exact DE

Example 1

(𝑠𝑖𝑛𝑦 − 𝑦 𝑠𝑖𝑛𝑥)𝑑𝑥 + (𝑐𝑜𝑠𝑥 + 𝑥𝑐𝑜𝑠𝑦 − 𝑦)𝑑𝑦 = 0 (iv)

Solution

Here 𝑀 = 𝑠𝑖𝑛𝑦 − 𝑦 𝑠𝑖𝑛𝑥 and 𝑁 = 𝑐𝑜𝑠𝑥 + 𝑥𝑐𝑜𝑠𝑦 − 𝑦

𝜕𝑀 𝜕𝑁
= 𝑐𝑜𝑠𝑦 − 𝑠𝑖𝑛𝑥 , = −𝑠𝑖𝑛𝑥 + 𝑐𝑜𝑠𝑦
𝜕𝑦 𝜕𝑥

𝜕𝑀 𝜕𝑁
We see that =
𝜕𝑦 𝜕𝑥

So the equation (iv) is exact.

𝜕𝑓 𝜕𝑓
Now let there be a function 𝑓(𝑥, 𝑦) = 𝑐 such that 𝑀 = and 𝑁 = .
𝜕𝑥 𝜕𝑦

5
 𝜕𝑓
Integrating 𝑀 = with respect to 𝑥, we have
𝜕𝑥

𝑓(𝑥, 𝑦) = 𝑥 𝑠𝑖𝑛𝑦 + 𝑦 𝑐𝑜𝑠𝑥 + 𝑔(𝑦) (v)

𝜕𝑓
So = 𝑥 𝑐𝑜𝑠𝑦 + 𝑐𝑜𝑠𝑥 + 𝑔′ (𝑦) = 𝑥 𝑐𝑜𝑠𝑦 + 𝑐𝑜𝑠𝑥 − 𝑦
𝜕𝑦

⇒ 𝑔′ (𝑦) = −𝑦

Integrating with respect to y , gives,

𝑦2
𝑔(𝑦) = −
2

Putting this value in (v) , we get

𝑦2
𝑓(𝑥, 𝑦) = 𝑥 𝑠𝑖𝑛𝑦 + 𝑦 𝑐𝑜𝑠𝑥 − = 𝑐 is the solution.
2

Example 2 (𝑡𝑎𝑛𝑥 − 𝑠𝑖𝑛𝑥 𝑠𝑖𝑛𝑦)𝑑𝑦 + 𝑐𝑜𝑠𝑥 𝑐𝑜𝑠𝑦 𝑑𝑦 = 0 So m (vi)

Solution Here 𝑀 = 𝑡𝑎𝑛𝑥 − 𝑠𝑖𝑛𝑥 𝑠𝑖𝑛𝑦 , 𝑁 = 𝑐𝑜𝑠𝑥 𝑐𝑜𝑠𝑦

𝜕𝑀 𝜕𝑁
Now = −𝑠𝑖𝑛𝑥 𝑐𝑜𝑠𝑦 , = −𝑠𝑖𝑛𝑥 𝑐𝑜𝑠𝑦
𝜕𝑦 𝜕𝑥

So equation (vi) is exact.

Then there is a function 𝑓(𝑥, 𝑦) = 𝑐 such that

𝜕𝑓 𝜕𝑓
= 𝑀, =𝑁
𝜕𝑥 𝜕𝑦

𝜕𝑓
Integrating = 𝑀 with respect to 𝑥,
𝜕𝑥

𝑓(𝑥, 𝑦) = ∫(𝑡𝑎𝑛𝑥 − 𝑠𝑖𝑛𝑥 𝑠𝑖𝑛𝑦)𝑑𝑥 + 𝑔(𝑦)

= 𝑙𝑛𝑠𝑒𝑐𝑥 + 𝑐𝑜𝑠𝑥 𝑠𝑖𝑛𝑦 + 𝑔(𝑦) (vii)

Now differentiating (vii) with respect to y,

6
 𝜕𝑓
𝑐𝑜𝑠𝑥 𝑐𝑜𝑠𝑦 + 𝑔′ (𝑦) = ⇒ 𝑐𝑜𝑠𝑥 𝑐𝑜𝑠𝑦 + 𝑔′ (𝑦) = 𝑁 = 𝑐𝑜𝑠𝑥 𝑐𝑜𝑠𝑦
𝜕𝑦

⇒ 𝑔′ (𝑦) = 0

Integrating

𝑔(𝑦) = 𝑘

Thus 𝑓(𝑥, 𝑦) = 𝑙𝑛𝑠𝑒𝑐𝑥 + 𝑐𝑜𝑠𝑥 𝑠𝑖𝑛𝑦 + 𝑘 = 𝑐 ⇒ 𝑙𝑛𝑠𝑒𝑐𝑥 + 𝑐𝑜𝑠𝑥 𝑠𝑖𝑛𝑦 = 𝑏 is the

solution.
3𝑦 2 −𝑡 2 𝑑𝑦 𝑡
Example 3 (
𝑦5
) 𝑑𝑡 + 2𝑦4 = 0 , 𝑦(1) = 1

Example 4 (4𝑡 3 𝑦 − 15𝑡 2−𝑦 )𝑑𝑡 + (𝑡 4 + 3𝑦 2 − 𝑡)𝑑𝑦 = 0.

Higher Order Linear ODEs

Examples

(1)
𝑑2𝑦 𝑑𝑦
2 −5 −3𝑦 =0 (i) , homogeneous equation.
𝑑𝑡 2 𝑑𝑡

Solution Let 𝑦 = 𝑒 𝑚𝑡 be a solution of (i).

𝑑𝑦 𝑑2𝑦
Then = 𝑚𝑒 𝑚𝑡 𝑎𝑛𝑑 = 𝑚2 𝑒 𝑚𝑡 .
𝑑𝑡 𝑑𝑡 2

Putting these into the equation (i) gives

2𝑚2 𝑒 𝑚𝑡 − 5𝑚𝑒 𝑚𝑡 − 3𝑒 𝑚𝑡 = 0 ⇒ (2𝑚2 − 5𝑚 − 3) 𝑒 𝑚𝑡 = 0

⇒ 2𝑚2 − 5𝑚 − 3 = 0 since 𝑒 𝑚𝑡 ≠ 0.

Solving 2𝑚2 − 5𝑚 − 3 = 0 (auxiliary equation) gives two value of 𝑚 ,

say 𝑚1 and 𝑚2 .

𝑚1 = 3, 𝑚2 = −1/2

7
 1
We get 𝑦1 = 𝑒 3𝑡 and 𝑦2 = 𝑒 −2𝑡 are independent/fundamental solutions.
1
The general solution is 𝑦(𝑡) = 𝐴𝑒 3𝑡 + 𝐵𝑒 −2𝑡 .

Now three cases arise:

Case (i) If 𝑚1 and 𝑚2 are real , distinct , then the general solution is

𝑦(𝑡) = 𝑐1 𝑒 𝑚1𝑡 + 𝑐2 𝑒 𝑚2𝑡 .

Case (ii) If 𝑚1 and 𝑚2 are real repeated that is 𝑚1 = 𝑚2 = 𝑚 , 𝑠𝑎𝑦,

then the solution is

𝑦(𝑡) = 𝑐1 𝑒 𝑚𝑡 + 𝑐2 𝑡𝑒 𝑚𝑡 = (𝑐1 + 𝑐2 𝑡)𝑒 𝑚𝑡 .

Case (iii) If 𝑚1 and 𝑚2 are complex, say 𝛼 ∓ 𝑖𝛽, then the solution

is 𝑦(𝑡) = 𝑒 𝛼𝑡 (𝐴 cos 𝛽𝑡 + 𝐵𝑠𝑖𝑛 𝛽𝑡).

Say, 𝑚 = −2 ∓ 3𝑖

Examples
𝑑2𝑦 𝑑𝑦
1. − 10 + 25𝑦 = 0
𝑑𝑡 2 𝑑𝑡

Solution We have the equation

𝑑2𝑦 𝑑𝑦
− 10 + 25𝑦 = 0 (ii)
𝑑𝑡 2 𝑑𝑡

Let 𝑦 = 𝑒 𝑚𝑡 be a solution of (ii) .

𝑑𝑦 𝑚𝑡
𝑑2𝑦
= 𝑚𝑒 , 2 = 𝑚2 𝑒 𝑚𝑡
𝑑𝑡 𝑑𝑡

8
Putting these into equation (ii) , we get

𝑚2 𝑒 𝑚𝑡 − 10𝑚𝑒 𝑚𝑡 + 25𝑒 𝑚𝑡 = 0 ⇒ (𝑚2 − 10𝑚 + 25)𝑒 𝑚𝑡 = 0

⇒ 𝑚2 − 10𝑚 + 25 since 𝑒 𝑚𝑡 ≠ 0.

⇒ (𝑚 − 5)2 = 0 ⇒ 𝑚 = 5,5

Then the general solution is 𝑦 = (𝐴 + 𝐵 𝑡)𝑒 5𝑡 .

𝑑2𝑥 𝑑𝑥
2. 2
+4 + 7𝑥 = 0 (iii)
𝑑𝑡 𝑑𝑡

Solution Take 𝑥 = 𝑒 𝑚𝑡 a solution of (iii)

𝑑𝑥 𝑑2𝑥
Then = 𝑚𝑒 𝑚𝑡 , = 𝑚2 𝑒 𝑚𝑡 .
𝑑𝑡 𝑑𝑡 2

Using these into (iii), we get

𝑚2 𝑒 𝑚𝑡 + 4 𝑚𝑒 𝑚𝑡 + 7𝑒 𝑚𝑡 = 0 ⇒ (𝑚2 + 4𝑚 + 7)𝑒 𝑚𝑡 = 0

⇒ 𝑚2 + 4𝑚 + 7 = 0 since 𝑒 𝑚𝑡 ≠ 0

−4 ∓ √16 − 28 −4 ∓ √−12 −4 ∓ 2𝑖 √3
⇒𝑚= = = = −2 ∓ 𝑖√3
2 2 2
Then the general solution is 𝑥 = 𝑒 −2𝑡 (𝐴𝑐𝑜𝑠 √3 𝑡 + 𝐵𝑠𝑖𝑛√3 𝑡).

𝑑2𝑥 𝑑𝑥
Example 3 4 +4 + 17𝑥 = 0, (𝑖𝑣) 𝑥(0) = −1, 𝑥 ′ (0) = 2 .
𝑑𝑡 2 𝑑𝑡

Solution Take 𝑥 = 𝑒 𝑚𝑡 a solution of (iv)

𝑑𝑥 𝑑2𝑥
Then = 𝑚𝑒 𝑚𝑡 , = 𝑚2 𝑒 𝑚𝑡
𝑑𝑡 𝑑𝑡 2

Using these into (iv), we get

4𝑚2 𝑒 𝑚𝑡 + 4 𝑚𝑒 𝑚𝑡 + 17𝑒 𝑚𝑡 = 0 ⇒ (4𝑚2 + 4𝑚 + 17)𝑒 𝑚𝑡 = 0

1
⇒ 4𝑚2 + 4𝑚 + 17 = 0 ⇒ 𝑚 = − ∓ 2 𝑖
2
9
 1
∴ 𝑥(𝑡) = 𝑒 −2𝑡 (𝐴 cos 2𝑡 + 𝐵𝑠𝑖𝑛 2𝑡) (v)
1 1
1
Now 𝑥 ′ (𝑡) = − 𝑒 −2𝑡 (𝐴 𝑐𝑜𝑠 2𝑡 + 𝐵𝑠𝑖𝑛 2𝑡) + 𝑒 −2𝑡 (−2𝐴 𝑠𝑖𝑛2𝑡 + 2𝐵 𝑐𝑜𝑠2𝑡)
2

…….. (vi)

Putting 𝑥(0) = −1, (v) gives

−1 = 𝐴

and putting 𝑥 ′ (0) = 2 , (vi) gives

1 1 1 1 3 3
2 = − ∙ 𝐴 + 2𝐵 ⇒ 2𝐵 = 2 + 𝐴 = 2 + (−1) = 2 − = ⇒ 𝐵 = .
2 2 2 2 2 4

Therefore the solution is

1
3
𝑥(𝑡) = 𝑒 −2𝑡 (− cos 2𝑡 + 𝑠𝑖𝑛 2𝑡).
4

𝑑3𝑦 𝑑2𝑦
Example 4 +3 − 4𝑦 = 0 (vii)
𝑑𝑡 3 𝑑𝑡 2

Solution Let 𝑦 = 𝑒 𝑚𝑡 be a solution of (vii)

𝑑𝑦 𝑑2𝑦 𝑑3𝑦
Then = 𝑚𝑒 𝑚𝑡 , = 𝑚2 𝑒 𝑚𝑡 , = 𝑚3 𝑒 𝑚𝑡
𝑑𝑡 𝑑𝑡 2 𝑑𝑡 3

(vii) becomes

𝑚3 𝑒 𝑚𝑡 + 3𝑚2 𝑒 𝑚𝑡 − 4𝑒 𝑚𝑡 = 0

⇒ (𝑚3 + 3𝑚2 − 4)𝑒 𝑚𝑡 = 0 ⇒ 𝑚3 + 3𝑚2 − 4 = 0

⇒ 𝑚2 (𝑚 − 1) + 4𝑚(𝑚 − 1) + 4(𝑚 − 1) = 0

⇒ (𝑚 − 1)(𝑚2 + 4𝑚 + 4) = 0 ⇒ (𝑚 − 1)(𝑚 + 2)2 = 0

⇒ 𝑚 = 1, 𝑚 = 2, 2.

The general solution is 𝑦 = 𝑐1 𝑒 𝑡 + (𝑐2 + 𝑐3 𝑡)𝑒 2𝑡 .

10
Exercise

11
Higher Order Nonhomogeneous Equations

To find a solution to equation of the form

𝑎𝑛 𝑦 (𝑛) + 𝑎𝑛−1 𝑦 (𝑛−1) + ⋯ + 𝑎1 𝑦 ′ + 𝑎0 𝑦 = 𝑔(𝑥) (i)

We must do two things

• find the complementary function 𝑦𝑐 and

• find any particular solution 𝑦𝑝 of the nonhomogeneous equation (i).

The general solution (i) is

𝑦 = 𝑦𝑐 + 𝑦𝑝 .
Here 𝑦𝑐 is called the complementary function which is the general solution of
the homogeneous part 𝑎𝑛 𝑦 (𝑛) + 𝑎𝑛−1 𝑦 (𝑛−1) + ⋯ + 𝑎1 𝑦 ′ + 𝑎0 𝑦 = 0
of equation (i).

12
 METHOD OF UNDETERMINED COEFFICIENTS

The first of two ways we shall consider for obtaining a particular solution 𝑦𝑝 for a
nonhomogeneous linear DE is called the method of undetermined coefficients. The
underlying idea behind this method is a conjecture about the form of 𝑦𝑝 , an
educated guess really, that is motivated by the kinds of functions that make up the
input function 𝑔(𝑥).

The general method is limited to linear DEs such as (i) where

• the coefficients 𝑎𝑖 , i=1,2,…n are constants and

• 𝑔(𝑥) is a constant 𝑘, a polynomial function, an exponential function𝑒 𝑎𝑥 , a sine

or cosine function 𝑠𝑖𝑛𝑏𝑥 or 𝑐𝑜𝑠 𝑏𝑥, or finite sums and products of these
functions.

Example 1
𝑑2𝑦 𝑑𝑦
+4 − 2𝑦 = 2𝑥 2 − 3𝑥 + 6 (ii)
𝑑𝑥 2 𝑑𝑥

Solution

Step 1 First we solve the associated homogeneous equation

𝑑2𝑦 𝑑𝑦
+4 − 2𝑦 = 0 (iii)
𝑑𝑥 2 𝑑𝑥

Take the solution to 𝑦 = 𝑒 𝑚𝑥 .

Then (iii) becomes

−4∓√16+8 −4∓2√6
𝑚2 + 4𝑚 − 2 = 0 ⇒ 𝑚 = = = −2 ∓ √6 , real roots .
2 2

Hence the complementary function is

𝑦𝑐 = 𝑐1 𝑒 (−2+√6)𝑥 + 𝑐2 𝑒 (−2−√6)𝑥

Step 2 Now since 𝑔(𝑥) = 2𝑥 2 − 3𝑥 + 6 is a quadratic polynomial let us

assume a particular solution of the form

13
𝑦𝑝 = 𝐴𝑥 2 + 𝐵𝑥 + 𝐶

We need to determine the unknown coefficients 𝐴, 𝐵, 𝐶.

𝑑𝑦𝑝 𝑑 2 𝑦𝑝
Now = 2𝐴𝑥 + 𝐵, = 2𝐴
𝑑𝑥 𝑑𝑥 2

Substituting these into the given equation (ii), we get

2𝐴 + 4(2𝐴𝑥 + 𝐵) − 2(𝐴𝑥 2 + 𝐵𝑥 + 𝐶) = 2𝑥 2 − 3𝑥 + 6

⇒ −2𝐴𝑥 2 + (8𝐴 − 2𝐵)𝑥 + 2𝐴 + 4𝐵 − 2𝐶 = 2𝑥 2 − 3𝑥 + 6 .

⇒ −2𝐴 = 2 ⇒ 𝐴 = −1
5
8𝐴 − 2𝐵 = −3 ⇒ 2𝐵 = −8 + 3 = −5 ⇒ 𝐵 = −
2
2𝐴 + 4𝐵 − 2𝐶 = 6 ⇒ −2 − 10 − 2𝐶 = 6 ⇒ 2𝐶 = −12 − 6 = −18 ⇒ 𝐶 = −9
5
∴ 𝑦𝑝 = −𝑥 2 − 𝑥 − 9
2
Step 3

Therefore the general solution is

5
𝑦 = 𝑦𝑐 + 𝑦𝑝 = 𝑐1 𝑒 (−2+√6)𝑥 + 𝑐2 𝑒 (−2−√6)𝑥 − 𝑥 2 − 𝑥 − 9.
2

Example 2
𝑑3𝑦 𝑑2𝑦
+ = 𝑒 𝑡 𝑐𝑜𝑠𝑡.
𝑑𝑡 3 𝑑𝑡 2

Solution

Step 1
𝑑3𝑦 𝑑2 𝑦
For homogeneous part + = 0 , let 𝑦 = 𝑒 𝑚𝑡 .
𝑑𝑡 3 𝑑𝑡 2

Then we have the auxiliary equation

𝑚3 + 𝑚2 = 0 ⇒ 𝑚 = 0, 0, −1.

14
So 𝑦𝑐 = 𝑐1 + 𝑐2 𝑡 + 𝑐3 𝑒 −𝑡 .

Step 2

To find the particular solution,

Let 𝑦𝑝 = 𝑒 𝑡 (𝐴 𝑐𝑜𝑠𝑡 + 𝐵𝑠𝑖𝑛𝑡).

𝑑𝑦𝑝
= 𝑒 𝑡 (𝐴 𝑐𝑜𝑠𝑡 + 𝐵𝑠𝑖𝑛𝑡) + 𝑒 𝑡 (−𝐴𝑠𝑖𝑛𝑡 + 𝐵𝑐𝑜𝑠𝑡) = 𝑒 𝑡 [(𝐴 + 𝐵)𝑐𝑜𝑠𝑡 + (𝐵 − 𝐴)𝑠𝑖𝑛𝑡)]
𝑑𝑥
𝑑 2 𝑦𝑝
= 𝑒 𝑡 [(𝐴 + 𝐵)𝑐𝑜𝑠𝑡 + (𝐵 − 𝐴)𝑠𝑖𝑛𝑡)] + 𝑒 𝑡 [−(𝐴 + 𝐵)𝑠𝑖𝑛𝑡 + (𝐵 − 𝐴)𝑐𝑜𝑠𝑡]
𝑑𝑥 2

= 2𝑒 𝑡 (𝐵𝑐𝑜𝑠𝑡 − 𝐴𝑠𝑖𝑛𝑡)
𝑑 3 𝑦𝑝
= 2𝑒 𝑡 (𝐵𝑐𝑜𝑠𝑡 − 𝐴𝑠𝑖𝑛𝑡) + 2𝑒 𝑡 [−𝐵𝑠𝑖𝑛𝑡 − 𝐴𝑐𝑜𝑠𝑡] = 2𝑒 𝑡 [(𝐵 − 𝐴)𝑐𝑜𝑠𝑡 −
𝑑𝑥 3
(𝐴 + 𝐵)𝑠𝑖𝑛𝑡].

Putting these into the given equation , we get

2𝑒 𝑡 [(𝐵 − 𝐴)𝑐𝑜𝑠𝑡 − (𝐴 + 𝐵)𝑠𝑖𝑛𝑡]+ 2𝑒 𝑡 (𝐵𝑐𝑜𝑠𝑡 − 𝐴𝑠𝑖𝑛𝑡) = 𝑒 𝑡 𝑐𝑜𝑠𝑡

⇒ 2(𝐵 − 𝐴) + 2𝐵 = 1 ⇒ 4𝐵 − 2𝐴 = 1

and

−2(𝐴 + 𝐵) − 2𝐴 = 0 ⇒ −4𝐴 − 2𝐵 = 0 ⇒ −2𝐴 = 𝐵

1
Then 4𝐵 − 2𝐴 = 1 ⇒ 4𝐵 + 𝐵 = 1 ⇒ 𝐵 =
5

𝐵 1 1
and 𝐴 = − = − =−
2 5.2 10

1 1
𝑦𝑝 = 𝑒 𝑡 [− 𝑐𝑜𝑠𝑡 + 𝑠𝑖𝑛𝑡]
10 5
Therefore the general solution is
1 1
𝑦 = 𝑦𝑐 + 𝑦𝑝 = 𝑐1 + 𝑐2 𝑡 + 𝑐3 e−t + 𝑒 𝑡 (− 𝑐𝑜𝑠𝑡 + 𝑠𝑖𝑛𝑡)
10 5

15
 𝑑2𝑥 𝑑𝑥
Example 3 Solve +4 + 4𝑥 = (3 + 𝑡)𝑒 −2𝑡 , 𝑥(0) = 2, 𝑥 ′ (0) = 5
𝑑𝑡 2 𝑑𝑡

𝑑2𝑥
Example 4 + 𝜔2 𝑥 = 𝐹0 𝑠𝑖𝑛𝜔𝑡 , 𝑥(0) = 0, 𝑥 ′ (0) = 0 .
𝑑𝑡 2

𝑑2 𝑥 𝑑𝑥
Example 5 +3 + 2𝑥 = 4 𝑡 2
𝑑𝑡 2 𝑑𝑡

Solution
𝑑2𝑥 𝑑𝑥
Step 1 We solve the homogeneous : +3 + 2𝑥 = 0
𝑑𝑡 2 𝑑𝑡

Let the solution be 𝑥 = 𝑒 𝑚𝑡

Then we have the auxiliary equation 𝑚2 + 3𝑚 + 2 = 0 ⇒ (𝑚 + 2)(𝑚 + 1) = 0

⇒ 𝑚 = −1, −2

So the complementary function is 𝑥𝑐 = 𝑐1 𝑒 −𝑡 + 𝑐2 𝑒 −2𝑡 .

Step 2

Let 𝑥𝑝 = 𝐴 + 𝐵𝑡 + 𝐶𝑡 2
𝑑𝑥𝑝 𝑑 2 𝑥𝑝
Then = 𝐵 + 2𝐶𝑡 and = 2𝐶
𝑑𝑡 𝑑𝑡 2

𝑑2𝑥 𝑑𝑥
Therefore +3 + 2𝑥 = 4 𝑡 2 gives
𝑑𝑡 2 𝑑𝑡

2𝐶 + 3(𝐵 + 2𝐶𝑡) + 2(𝐴 + 𝐵𝑡 + 𝐶𝑡 2 ) = 4 𝑡 2

⇒ 2𝐶 = 4 ⇒ 𝐶 = 2

6𝐶 + 2𝐵 = 0 ⇒ 2𝐵 = −6𝐶 ⇒ 𝐵 = −3𝐶 = −6

2𝐶 + 3𝐵 + 2𝐴 = 0 ⇒ 4 − 18 + 2𝐴 = 0 ⇒ 2𝐴 = 14 ⇒ 𝐴 = 7

Thus 𝑥𝑝 = 𝐴 + 𝐵𝑡 + 𝐶𝑡 2 = 7 − 6𝑡 + 2𝑡 2

The general solution is = 𝑥𝑐 + 𝑥𝑝 = 𝑐1 𝑒 −𝑡 + 𝑐2 𝑒 −2𝑡 + 7 − 6𝑡 + 2𝑡 2 .

16
 𝑑2 𝑥 𝑑𝑥
Example 6 Find the solution of − 2 𝑑𝑡 + 𝑥 = 𝑒 𝑡
𝑑𝑡 2

Solution

Step 1 Let the solution be 𝑥 = 𝑒 𝑚𝑡 .

Then we have the auxiliary equation 𝑚2 − 2𝑚 + 1 = 0 ⇒ 𝑚 = 1,1

So 𝑥𝑐 = 𝑐1 𝑒 𝑡 + 𝑐2 𝑡 𝑒 𝑡

Step 2

Here 𝑥𝑝 = 𝐴𝑒 𝑡 will duplicate the solution in 𝑥𝑐 .

Also 𝑥𝑝 = 𝐴𝑡𝑒 𝑡 will again duplicate the solution in 𝑥𝑐 .

Then let 𝑥𝑝 = 𝐴 𝑡 2 𝑒 𝑡

𝑑𝑥𝑝
∴ = 2𝐴𝑡𝑒 𝑡 + 𝐴 𝑡 2 𝑒 𝑡
𝑑𝑡
𝑑 2 𝑥𝑝
2
= 2𝐴𝑒 𝑡 + 2𝐴 𝑡𝑒 𝑡 + 2𝐴𝑡𝑒 𝑡 + 𝐴 𝑡 2 𝑒 𝑡 = 2𝐴𝑒 𝑡 + 4𝐴𝑡𝑒 𝑡 + 𝐴 𝑡 2 𝑒 𝑡
𝑑𝑡
𝑑2𝑥 𝑑𝑥
Thus 2
−2 + 𝑥 = 𝑒 𝑡 becomes
𝑑𝑡 𝑑𝑡

2𝐴𝑒 𝑡 + 4𝐴𝑡𝑒 𝑡 + 𝐴 𝑡 2 𝑒 𝑡 − 2(2𝐴𝑡𝑒 𝑡 + 𝐴 𝑡 2 𝑒 𝑡 ) + 𝐴 𝑡 2 𝑒 𝑡 = 𝑒 𝑡

1
⇒ 2𝐴 = 1 ⇒ 𝐴 =
2
1
So 𝑥𝑝 = 𝑡 2𝑒 𝑡 .
2

Exercise 4.4

17
18