Differential Equations :D

xp
17 December 2020

1
1 The general solution of first-order ODEs
A first order, linear differential equation is one of the following form:
dy
+ Py = Q (1.1)
dx
We can find a general solution to the above as follows.
Thm. 1. The general solution of a first order, linear differential equation in
the form as shown in (1.1) is
Z
1
y= IQ dx (1.2)
I
where I is defined as R
I = e P dx (1.3)
Proof.
dy
+ Py = Q
dx
We proceed by multiplying both sides by an integrating factor, I.
dy
I + IP y = IQ
dx
Let
dI
IP = (1.4)
dx
and hence
dy dI
I + y = IQ
dx dx
d
(Iy) = IQ
dx Z
Iy = IQ dx
Z
1
y= IQ dx (1.5)
I
However, we have yet to define I. From (1.4):
1 dI
=P
Z I dx Z
1
dI = P dx
I
Z
ln |I| = P dx + c
R R
P dx+c P dx
I=e = Ae (A = ec )

2
We can disregard the arbitrary constant A as it will always be a multiplier
and never be 0 (ec 6= 0 ∀c). Hence
R
P dx
I=e (1.6)
Equations (1.5) and (1.6) prove Thm. 1.
QED

2 The general solution of second-order homo-

geneous ODEs
A second order, homogeneous, linear differential equation is one of the fol-
lowing form:
d2 y dy
2
+b + cy = 0
dx dx
We can find a general solution to the above as follows.
Thm. 2. Given a second order, homogeneous, linear differential equation in
the following form:
d2 y dy
2
+b + cy = 0 (2.1)
dx dx
and α and β are defined as the roots of the equation λ2 + bλ + c = 0, the
general solution of the differential equation as shown in (2.1) is as follows.
Case 1. α 6= β
y = Aeαx + Beβx (2.2)
Case 2. α = β
y = Aeαx + Bxeαx = (A + Bx)eαx (2.3)
where in both cases, A and B are arbitrary constants.
Proof. As α and β are roots of λ2 + bλ + c = 0
λ2 + bλ + c = 0 =⇒ λ2 − (α + β)λ + αβ
=⇒ b = −(α + β)
=⇒ c = αβ
Therefore (2.1) can be rewritten as:
d2 y dy
− (α + β) + αβy = 0
dx2 dx
d2 y dy dy
2
−α −β + αβy = 0
dx
 dx dx
 
d dy dy
− αy − β − αy = 0 (2.4)
dx dx dx

3
Now let
dy
u= − αy (2.5)
dx
and hence, from (2.4)
du
− βu = 0
dx
This is a first order, linear ODE which we can solve by Thm. 1. Therefore,
where B is an arbitrary constant,
u = Beβx
We can then substitute this into (2.5) and therefore:
dy
− αy = Beβx (2.6)
dx
Finally, we can solve this by Thm. 1.
Z
1
y = −αx e−αx Beβx dx
e
Z
αx
=e Be(β−α)x dx (2.7)

We observe that our integral is evaluated differently depending on whether

β − α is zero or non-zero. Hence we proceed considering the two cases sepa-
rately.
Case 1. α = β (λ2 + bλ + c = 0 has repeated roots.)
Z
αx
y=e Be(β−α)x dx
Z
αx
=e Be0 dx

= eαx (Bx + A) (2.8)

Case 2. α 6= β (λ2 + bλ + c = 0 has two distinct roots.)
Z
αx
y=e Be(β−α)x dx
 
αx B (β−α)x
=e e +A
β−α
= eαx Be(β−α)x + A

= Aeαx + Beβx (2.9)

(2.8) and (2.9) prove Thm. 2.
QED

4
3 The general solution of second-order non-
homogeneous ODES
A second order, non-homogeneous, linear differential equation is similar to
its homogeneous counterpart and is represented in following form:

d2 y dy
2
+b + cy = f (x)
dx dx
We can find a general solution to the above with a similar method as previ-
ously as follows.

Thm. 3. Given a second order, non-homogeneous, linear differential equa-

tion in the following form:

d2 y dy
+ b + cy = f (x) (3.1)
dx2 dx
and α and β are defined as the roots of the equation λ2 + bλ + c = 0, the
general solution of the differential equation as shown in (3.1) is as follows.
Z  Z 
αx (β−α)x −βx
y=e e e f (x) dx dx (3.2)

Proof. It can be shown that b = −(α + β) and c = αβ.

d2 y dy
2
− (α + β) + αβy = f (x)
dx dx
d2 y dy dy
2
−α −β + αβy = f (x)
dx
 dx dx
 
d dy dy
− αy − β − αy = f (x) (3.3)
dx dx dx

Now let
dy
u= − αy (3.4)
dx
and hence, from (3.3)
du
− βu = f (x)
dx
By Thm. 1, we can solve for u.
Z
u=e βx
e−βx f (x) dx

5
By application of Thm. 1. on (3.4), we can solve for y.
Z
y=e αx
e−αx u dx
Z  Z 
αx −αx βx −βx
=e e e e f (x) dx dx
Z  Z 
αx (β−α)x −βx
=e e e f (x) dx dx (3.5)

(3.5) directly proves Thm. 3.

QED

4 An example: a linear function

Thm. 4. Given a second order, non-homogeneous, linear differential equa-
tion in the following form:
d2 y dy
2
+b + cy = ax + b (4.1)
dx dx
and α and β are defined as the roots of the equation λ2 + bλ + c = 0, the
general solution of the differential equation as shown in (4.1) is as follows.
 
αx βx a a a + βb
y = Ae + Be + x− 2 + (4.2)
αβ α β αβ 2
Proof. Let f (x) = ax + b. By Thm. 3.,
Z Z 
αx (β−α)x −βx
y=e e e (ax + b) dx dx

We begin by dealing with each of the nested integrals. Let A and B be

arbitrary constants.
Z Z  
−βx −βx βx −βx aβx + βb + a
e (ax + b) dx = axe + be dx = −e +B
β2
Z Z  Z    
(β−α)x −βx (β−α)x aβx + βb + a
−βx
e e (ax + b) dx dx = − e −e + B dx
β2
Z  
−αx aβx + βb + a
=− e + Be(β−α)x dx
β2
   
a x −αx 1 −αx a + βb −αx
=− − e − 2e + e
β a α aβ 2
+Be(β−α)x + A

6
Finally, we can calculate y.
     
αx a x −αx 1 −αx a + βb −αx (β−α)x
y=e − − e − 2e + e + Be +A
β a α aβ 2
ax a a + βb
= − 2 + 2
+ Beβx + Aeαx
αβ α β αβ
 
αx βx a a a + βb
= Ae + Be + x− 2 + (4.3)
αβ α β αβ 2
(4.3) directly proves Thm. 4.
QED

5 Trial functions
We utilise trial functions to easily find the general solution to specific non-
homogeneous ODEs. For example, by Thm. 4., we can consider the case
where the arbitrary constants are equal to zero, and hence a solution to the
ODE is  
a a a + βb
y= x− 2 + (5.1)
αβ α β αβ 2
We recognise this is in the form of y = ax + b, and hence, we can substitute
this into the ODE and equate coefficients in order to find a and b.

6 Appendices
A Alternative form for non-real roots
An alternative form for when the auxiliary equation does not have any real
roots is shown as follows.
α, β ∈/ R. Let α = a + bi and β = α∗ = a − bi (a, b ∈ R) and hence
y = Aeαx + Beβx
= Ae(a+bi)x + Be(a−bi)x
= eax Aeibx + Be−ibx

= eax (A (cos(bx) + i sin(bx)) + B (cos(bx) − i sin(bx))))

= eax ((A + B) cos(bx) + i(A − B) sin(bx))
Finally, let P = (A + B) and Q = i(A − B) as arbitrary constants.
y = eax (P cos(bx) + Q sin(bx))