1 1 1

1. sin x = sec x = tan x =

csc x cos x cot x
1 1 1
csc x = cos x = cot x =
sin x sec x tan x
sin x cos x
tan x = cot x =
cos x sin x
2. sin 2 x + cos 2 x = 1
1 + tan 2 x = sec 2 x 1 + cot 2 x = csc 2 x
π  π  π 
3. sin − x  = cos x cos − x  = sin x tan − x  = cot x
2  2  2 
π  π  π 
csc − x  = sec x sec − x  = csc x cot  − x  = tan x
2  2  2 
4. sin ( − x ) = − sin x cos( − x ) = cos x
csc( − x ) = − csc x tan ( − x ) = − tan x
sec( − x ) = sec x cot ( − x ) = − cot x
5. sin ( u ± v ) = sin u cos v ± cos u sin v
cos( u ± v ) = cos u cos v sin u sin v
tan u ± tan v
tan ( u ± v ) =
1 tan u tan v
6. sin 2u = 2 sin u cos u
cos 2u = cos 2 u − sin 2 u = 2 cos 2 u − 1 = 1 − 2 sin 2 u
2 tan u
tan 2u =
1 − tan 2 u
1 − cos 2u
7. sin 2 u =
2
1 + cos 2u
cos 2 u =
2
1 − cos 2u
tan 2 u =
1 + cos 2u
1
8. sin u sin v = [ cos( u − v ) − cos( u + v ) ]
2
1
cos u cos v = [ cos( u − v ) + cos( u + v ) ]
2
1
sin u cos v = [ sin( u + v ) + sin( u − v ) ]
2
1
cos u sin v = [ sin( u + v ) − sin( u − v ) ]
2

1
9. a 0 = 1, a ≠ 0 ( ab ) x = a xbx a x a y = a x+ y a = a1 2

x
ax a ax
= a x− y n
a =a 1n
  = x n
am = am n
ay b b

1 a na
a −x =
ax
n
ab = n a n b (a )
x y
=a xy n =
b nb
x
10. (a) log a ( xy ) = log a x + log a y (b) log a  y  = log a x − log a y
 

(c) ( )
log a x r = r log a x

e x − e−x e x + e− x sinh x
11. (a) sinh x = (b) cosh x = (c) tanh x =
2 2 cosh x
1 1 cosh x
(d) csch x = sinh x (e) sech x = cosh x (f) coth x =
sinh x
12. (a) sinh ( − x ) = − sinh x (b) cosh ( − x ) = cosh x
(c) cosh 2 x − sinh 2 x = 1 (d) 1 − tanh x = sech x
2 2

(f) sinh ( x + y ) = sinh x cosh y + cosh x sinh y

(g) cosh ( x + y ) = cosh x cosh y + sinh x sinh y
d d
13. (a) ( sinh x ) = cosh x (b) ( csch x ) = −csch x coth x
dx dx
d d
(c) ( cosh x ) = sinh x (d) ( sech x ) = −sech x tanh x
dx dx
d d
(e) ( tanh x ) = sech2 x (f) ( coth x ) = −csch 2 x
dx dx

14. (a)
d
(
sinh −1 x =) 1
(b)
d
dx
(
csch −1 x = − )
1
dx 1 + x2 x x2 + 1

(c)
d
dx
(
cosh −1 x =) 1
(d)
d
dx
(
sech −1 x = − )1
x2 − 1 x 1 − x2

(e)
d
dx
(
tanh −1 x =) 1
1 − x2 (f)
d
dx
(
coth −1 x =
1
)
1 − x2
d d
15. dx [ cu ] = cu ' 16. dx [ u ± v ] = u′ ± v′

d d  u  vu′ − uv′
17. dx [ uv ] = uv′ + vu′ 18. dx  v  = v2
 
d
19. dx [ c ] = 0
d n
[ ]
20. dx u = nu u′
n −1

2
 d
21. dx [ x ] = 1
d
22. dx [ u ] =
u
u
( u′) , u ≠ 0
u′
d
23. dx [ ln u ] = u
d u
[ ]
24. dx e = e u′
u

d u
[ ] (
25. dx a = a ln a u′
u
) d  1 
26. dx [ log a u ] =  u ln a  u′
 
d d
27. dx [ sin u ] = ( cos u ) u′ 28. dx [ cos u ] = −( sin u ) u′

d
(
29. dx [ tan u ] = sec u u′
2
) d
(
30. dx [ cot u ] = − csc u u′
2
)
d d
31. dx [ sec u ] = ( sec u tan u ) u′ 32. dx [ csc u ] = −( csc u cot u ) u′
d u′ d − u′
33. dx [ arcsin u ] = 34. dx [ arccos u ] =
1 − u2 1− u2
d u′ d − u′
35. dx [ arctan u ] = 1 + u 2 36. dx [ arccot u ] = 1 + u 2

d u′ d − u′
37. dx
[ arcsec u ] =
38. dx
[ arccsc u ] =
u u2 −1 u u2 −1

u can be any function of x .

d
[
Ex. dx sin e
x
]
Let u = e x

d x
= cos e x e
dx
= e x cos e x

39. ∫ kf ( u ) du = k ∫ f ( u ) du 40. ∫ [ f ( u ) ± g ( u ) ] du = ∫ f ( u ) du ± ∫ g ( u ) du
u n +1
41. ∫ du = u + C 42. ∫ u du = + C , n ≠ −1
n

n +1
du
∫ ∫ e du = e +C
u u
43. = ln u + C 44.
u

45. ∫ sin u du = − cos u + C 46. ∫ cos u du = sin u + C

47. ∫ tan u du = − ln cos u + C 48. ∫ cot u du = ln sin u + C

49. ∫ sec u du = ln sec u + tan u + C 50. ∫ csc u du = − ln csc u + cot u + C

3
 ∫ sec u du = tan u + C ∫ csc u du = − cot u + C
2 2
51. 52.

53. ∫ sec u tan u du = sec u + C 54. ∫ csc u cot u du = − csc u + C

du u du 1 u
55. ∫ a −u
2 2
= arcsin
a
+C 56. ∫a 2
+u 2
= arctan + C
a a
du 1 u
57. ∫u u2 − a2
=
a
arcsec + C
a
58. The Substitution Rule
If u = g ( x )

∫ f ( g ( x ) ) g ′( x ) dx = ∫ f ( u ) du

Ex. Find ∫x
3
(
cos x 4 + 2 dx )
Let u = x4 + 2
Then du = 4 x 3dx
Thus, using x 3dx = du 4 and the Substitution Rule, we have

∫x
3
( )
cos x 4 + 2 dx = ∫ cos u ⋅
1
4
1
du = ∫ cos u du
4
1
= sin u + C
4
1
( )
= sin x 4 + 2 + C
4

59. Formula for integration by parts

∫ u dv = uv − ∫ v du

∫x
m ax
Ex. Find e dx

Let u = xm dv = e ax dx
1 ax
Then du = mx m −1dx v= e
a

∫x
m ax
and so e dx

1 m ax 1
= x e − ∫ e ax mx m −1dx
a a
1 m
= x m e ax − ∫ x m −1e ax dx
a a

4
60. Euler’s formula
eibt = cos bt + i sin bt

Ex. ∫e e dt = ∫ e at ( cos bt + i sin bt ) dt

at ibt

= ∫ e at cos bt dt + i ∫ e at sin bt dt

1
∫e e dt = ∫ e( a + ib ) t dt =
at ibt
e( a + ib ) t
a + ib
a − ib
= e at eibt
( a + ib )( a − ib )
a − ib at
= 2 e ( cos bt + i sin bt )
a + b2
 a b 
= 2 e at cos bt + 2 e at sin bt 
a +b a +b
2 2

 a b 
+ i 2 e at sin bt − 2 e at cos bt 
a +b a +b
2 2

and so
 a b 
∫e cos bt dt = e at  2 cos bt + 2
at
real part: sin bt 
a +b a +b
2 2

at  a b 
∫ e sin bt dt = e  a 2 + b 2 sin bt − a 2 + b 2 cos bt 
at
imaginary part:

61. Rational functions in integration or Laplace transform

R( x )
Q( s ) deg ( Q ) > deg ( R )

f 2 − 4eg < 0
m 2 − 4hn < 0
R( s )
( ax + b )( cx + d ) 2
( ex
+ fx + g hx 2 + mx + n
2
)( ) 2

A1 A2 A3 A x + A5 A6 x + A7 A8 x + A9
= + + + 24 + +
(
( ax + b ) ( cx + d ) ( cx + d ) ex + fx + g hx + mx + n hx + mx + n 2
2 2 2 ) ( ) ( )
Determine A , A ,, A and then integrate or find the inverse Laplace transform
1 2 9

5
62. Change to Polar Coordinates in a Double Integral

∫ ∫ f ( x, y ) dA = ∫ ∫ f ( r cosθ , r sin θ ) r dr dθ
where x = r cos θ
y = r sin θ

63. Formula for Triple Integration in Cylindrical Coordinates

∫ ∫ ∫ f ( x, y, z ) dV = ∫ ∫ ∫ f ( r cosθ , r sin θ , z ) r dz dr dθ
where x = r cosθ
y = r sin θ
z=z
64. Formula for Triple Integration in Spherical Coordinates

∫ ∫ ∫ f ( x, y, z ) dV = ∫ ∫ ∫ f ( ρ sin φ cosθ , ρ sin φ sin θ , ρ cosφ ) ρ sin φ dρ dθ dφ

2

where x = ρ sin φ cosθ

y = ρ sin φ sin θ
z = ρ cos φ

6
Chapter 1 Ordinary Differential Equations
1.1 Preliminary Concepts
若 方程式數目=m; 變數數目=n
自由度：獨立變數數目 k=n-m
一階常微分方程式之獨立變數數目為 1

F ( x, y , y ′ ) = 0 => y:非獨立變數

x:獨立變數
G ( x, y , y ′ ) = 0 => x:非獨立變數

y:獨立變數

方程式之解：將非獨立變數表示為獨立變數之函數

y = F ( x ) 或 x = g ( y ) 或 h ( x, y ) = 0

Initial value problem

F ( x, y , y ′ ) = 0 y ( x0 ) = y0

1.2 Separable Equations

利用指數、對數、積化和差及因數分解將微分方程式化成

A( x ) dx = B( y ) dy

再分別積分.

Ex: [ cos( x + y ) + sin ( x − y ) ] dy = cos( 2 x )

dx

7
 cos( x + y ) + sin ( x − y )
= cos x cos y − sin x sin y + sin x cos y − cos x sin y
= cos x( cos y − sin y ) + sin x( cos y − sin y )
= ( cos x + sin x )( cos y − sin y )
cos 2 x = cos 2 x − sin 2 x = ( cos x + sin x )( cos x − sin x )

[ cos( x + y ) + sin ( x − y ) ] dy = cos( 2 x )

dx
⇒ cos y − sin y dy = cos x − sin x ) dx
( ) (
⇒ sin y + cos y = sin x + cos x + C
1.3 Linear differential equations 線性
獨立變數及其微分式均為一次式
A first-order differential equation is linear if it has the form

y′( x ) + p( x ) y = q( x )

p ( x ) dx
Multiply the differential equation by e ∫ to get

p ( x ) dx p ( x ) dx p ( x ) dx
e∫ y′( x ) + p( x ) e ∫ y = q( x ) e ∫
d  ∫ p ( x ) dx  = q( x ) e ∫ p ( x ) dx
 y( x ) e 
dx  
= ∫  q( x ) e ∫  dx + C
p ( x ) dx p ( x ) dx
y( x ) e∫ 
 
− p ( x ) dx  ∫ p ( x ) dx  dx + Ce − ∫ p ( x ) dx
y( x ) = e ∫ ∫  q( x ) e 


Ex: sin ( 2 x ) y′ + 2 y sin 2 x = 2 sin x

8
 2 sin x cos xy′ + 2 y sin 2 x = 2 sin x
sin x 1
y′ + y=
cos x cos x
sin x 1
∫ cos x dx − ∫ cos x d ( cos x ) 1
= e − ln cos x = eln ( cos x ) =
−1
e =e
cos x
1 sin x 1
y′ + 2
y=
cos x cos x cos 2 x
′
 1  1
 y = 2
 cos x  cos x
1 1
y=∫ 2
dx = ∫ sec 2 x dx = tan x + C
cos x cos x
 sin x 
y = cos x + c  = sin x + C cos x
 cos x 

1.4 Exact differential equations

Mdx + Ndy = 0

Theorem 1.1 Test for Exactness

Suppose M ( x, y ) , N ( x, y ) , ∂M ∂y , and ∂N ∂x are continuous for all ( x, y )
within a rectangle R in the plane. Then,
M ( x, y ) + N ( x, y ) y ′ = 0
is exact on R if and only if, for each ( x, y ) in R,

∂M ∂N
=
∂y ∂x

Exact 解為一 potential function Φ ( x, y ) = C

2 x 3 + 2 xy 2 − y x
Ex: + 2 y′ = 0
x +y
2 2
x + y2
2 x 3 + 2 xy 2 − y x
dx + 2 dy = 0
x +y
2 2
x + y2
 2 x 3  2 xy 2  y  x
M   N 
x x2  y 2 y x  y2
2

M ( x 2  y 2 )(4 xy  1)  (2 x 3  2 xy 2  y )(2 y )

y ( x2  y 2 )2

9
 4 x 3 y  x 2  4 xy 3  y 2  4 x 3 y  4 xy 3  2 y

( x 2  y 2 )2
y 2  x2
 2
( x  y 2 )2
N ( x 2  y 2 )  x(2 x) y 2  x2
  2
x ( x 2  y 2 )2 ( x  y 2 )2
 x
 2
y x  y 2
x
 dy  h( x )
x  y2
2

1 1 1 y
 
x 1  ( y )2
dy  h( x)   d  h( x )
y 2 x
y
 tan 1 ( )  h( x)
1 ( ) x
x x
 1 y
 ( 2 ) y
x 1  ( ) 2 x  2
y  h( x)
x  y2
x
2 x( x 2  y 2 )
h( x)   2x
x2  y 2
h( x )  x 2
y
tan 1 ( )  x 2  C #
x
1.5 Integrating Factors
If ( µM ) dx + ( µN ) dy = 0 is exact. Therefore, we must have

∂
[ µM ] = ∂ [ µN ]
∂y ∂x

Then
∂µ ∂M ∂µ ∂N
M +µ =N +µ
∂y ∂y ∂x ∂x
∂µ
µ = µ ( x ) , then =0
∂y
∂M ∂N dµ
µ =µ +N
∂y ∂x dx
1 dµ 1  ∂M ∂N 
=  − 
µ dx N  ∂y ∂x 

Then equation is a separable differential equation for an integrating factor µ ( x )

10
If µ = µ( y)
1 dµ 1  ∂N ∂M 
Then =  − 
µ dy M  ∂x ∂y 

Ex: 6 xy + 2 y + 8 + xy′ = 0
(6 xy + 2 y + 8)dx + xdy = 0
∂M ∂N
− = 6x + 2 −1 = 6x + 1
∂y ∂x
1 ∂M ∂N 6x +1 1
( − )= = 6+
N ∂y ∂x x x
1 du 1 du 1
= 6+ = (6 + )dx ln u  6 x  ln x  C
u dx x u x
u u
ln  6x  C  e6 x  c  C1e6 x
x x

u  C1 xe6 x C1  1  u  xe6 x

xe6 x (6 xy  2 y  8) dx  x 2 e6 x dy  0

   x 2 e6 x dy  x 2 e6 x y  k ( x)


 2 xe6 x y  6 x 2 e6 x y  k ( x)  6 x 2 e6 x y  2 xye6 x  8 xe6 x
x
 k ( x )  8 xe6 x
1 1
 xe dx   xde6 x  [ xe6 x   e6 x dx]
6x

6 6
1 6x 1 6x 4 2
 xe  e  k ( x)  xe6 x  e6 x
6 36 3 9
4 6x 2 6x
 x 2 e6 x y  xe  e  C #
3 9

1.6 Homogeneous, Bernoulli, and Riccati Equations

Homogeneous Equation
A first-order different equation is homogeneous if it has the form
 y
y  f  
 x

y  ux

11
 To see this, compute y′ = u′x + x′u = u′x + u and write u = y x . Then
y′ = f ( y x ) becomes
u′x + u = f ( u )
We can write this as
1 du 1 1 1
= or du = dx
f ( u ) − u dx x f (u) − u x
The Bernoulli Equation
A Bernoulli equation is a first-order equation,
y′ + P ( x ) y = R ( x ) yα
in which α is a real number
令 v = y1−α

代替原方程式，將其轉換為以 v 為非獨立變數之線性方程式求解

The Riccati Equations

A different equation of the form
y′ = P ( x ) y 2 + Q ( x ) y + R ( x )
is called a Riccati equation.
1
令 y = S ( x) + z

代替原方程式，將其轉換為以 z 為非獨立變數之線性方程式求解

2x − 5 y − 9
Ex: y′ = − 4 x + y + 9
2 x − 5 y − 9 = 0  x  2 
解   =  
− 4 x + y + 9 = 0  y  − 1
Χ = x + γ Χ = 0 x = 2 γ = −2 Χ = x − 2 x = Χ + 2
 
Υ = y + q Υ = 0 y = −1 γ = 1 Υ = y + 1 y = Υ − 1

12
 dΥ d ( y + 1) dy
= = = y′
dΧ d ( x − 2) dx
Υ
2−5
dΥ 2 Χ − 5Υ Χ , u = Υ , Υ = uΧ, Υ′ = u′Χ + u
= Υ′ = =
dΧ − 4Χ + Υ − 4 + Υ Χ
Χ
2 − 5u
u′Χ + u =
−4+u
2 − 5u + 4u − u 2
u′Χ =
−4+u
du 1 2 − u − u 2
=
dΧ Χ − 4 + u
4−u 1
∫ u +u −2
2
du = ∫ Χ dΧ + C
1 2
∫ u − 1 du − ∫ u + 2 du = ln Χ + C
ln ( u − 1) − 2 ln( u + 2 ) = ln Χ + C
u −1
ln = ln Χ + C
( u + 2) 2
u −1
= CΧ
( u + 2) 2
Υ
−1
Χ = CΧ
2
Υ 
 + 2
Χ 
1
( Υ − Χ)
Χ = CΧ
1
( Υ + 2Χ ) 2

Χ2
( Υ − Χ ) = C ⇒ ( y + 1 − x + 2) = C
( Υ + 2Χ ) 2 ( y + 1 + 2( x − 2) ) 2
⇒
( y − x + 3) = C
( y + 2 x − 3) 2

x − 2y
Ex: y′ = 3 x − 6 y + 4

13
 x u 1 1
u = x − 2y y= − y′ = − u′
2 2 2 2
1 1 u
− u′ =
2 2 3u + 4
2u u+4
u′ = 1 − =
3u + 4 3u + 4
du u + 4
=
dx 3u + 4
3u + 4
∫ u + 4 du = ∫ dx + C
8
∫ 3 − u + 4 du = x + C′
1
∫ 3 du − 8∫ u + 4 d ( u + 4) = x + C′
3u − 8 ln ( u + 4 ) = x + C ′
3( x − 2 y ) − 8 ln ( x − 2 y + 4 ) = x + C ′
2 x − 6 y − 8 ln ( x − 2 y + 4 ) = C ′
x − 3 y − 4 ln ( x − 2 y + 4 ) = C ′′

CHAPTER 2 Second-Order Differential Equations

Let y1 and y2 be solutions of y ′′ + p( x ) y′ + q( x ) y = 0 on an open interval I. Then,
y1 ( x ) y2 ( x ) y ( x) y2 ( x )
1. Either y′( x ) = 0 for all x in I, or 1 ≠ 0 for all x in I.
1 y2′ ( x ) y1′( x ) y′2 ( x )

y1 ( x ) y2 ( x )
y y
2. 1 and 2 are linearly independent on I if and only if y′( x ) ≠ 0 on I.
1 y′2 ( x )

Let y1 and y2 be solutions of y ′′ + p( x ) y′ + q( x ) y = 0 on an open interval I.

1. y1 and y2 form a fundamental set of solutions on I if y1 and y2 are linear
independent on I.
2. When y1 and y2 form a fundamental set of solutions, we call c1 y1 + c2 y 2 , with
c1 and c2 arbitrary constants, the general solution of the differential equation on
I.

The Homogeneous Equation y ′′ + p( x ) y′ + q( x ) y = 0

The Nonhomogeneous Equation y′′ + p ( x ) y′ + q( x ) y = f ( x )

This conclusion leads us to call c1 y1 + c2 y2 + y p the general solution of the

14
nonhomogeneous equation suggests the following strategy.
To solve y′′ + p ( x ) y′ + q( x ) y = f ( x )
1. find the general solution c1 y1 + c2 y 2 of the associated homogeneous equation
y ′′ + p( x ) y′ + q( x ) y = 0

2. find any solution y p of y′′ + p ( x ) y′ + q( x ) y = f ( x ) , and

3. write the general solution c1 y1 + c2 y2 + y p . This expressions contains all possible

solutions of y′′ + p ( x ) y′ + q( x ) y = f ( x ) on the interval

Reduction of Order
2 y1′( x ) + p( x ) y1 ( x )
y ( x ) = u ( x ) y1 ( x ) g ( x) =
y1 ( x )
y2′ = u′y1 + uy1′, y′′ = u′′y1 + 2u′y1′ + uy1′′
u′′ + g ( x ) u′ = 0
u′′y1 + 2u′y1′ + uy1′′ + p[ u′y1 + uy′] + quy1 = 0
v′ + g ( x ) v = 0
u′′y1 + u′[ 2 y1′ + py1 ] + u[ y1′′ + py′ + qy1 ] = 0 − g ( x ) dx
u′′y1 + u′[ 2 y1′ + py1 ] = 0 v( x ) = Ce ∫
− g ( x ) dx
2 y′ + py1 v( x ) = e ∫
u′′ + 1 u′ = 0
y1 − g ( x ) dx
u( x) = ∫ e ∫ dx

The Constant Coefficient Homogeneous Linear Equation

y′′ + Ay′ + By = 0
Case 1: A − 4 B > 0
2

y ( x ) = c1eax + c2ebx
Case 2: A − 4 B = 0
2

y ( x ) = e − Ax 2 ( c1 + c2 x )
Case 3: A − 4 B < 0
2

y ( x ) = c1e( p + iq ) x + c2e( p −iq ) x

An Alternative General Solution in the Complex Root Case
y ( x ) = e px ( c1 cos( qx ) + c2 sin ( qx ) )

Ex: y′′ = 1 + ( y′)

2

15
 u = y′
u′ = 1 + u 2
1
∫ 1 + u 2 du = ∫ dx
tan −1 u = x + C1
tan −1 y′ = x + C1
y ′ = tan ( x + C1 )
− d cos( x + C1 )
∫ dy =∫ tan( x + C ) dx = ∫
1
cos( x + C1 )
= − ln cos( x + C1 )

y = ln sec( x + C1 ) + C2

Ex: yy′′ = y y′ + ( y′)

2 2

d  dy  du du dy du
u = y′ y ′′ =  = = =u
dx  dx  dx dy dx dy
du
yu = y 2u + u 2
dy
du
y = y2 + u
dy
1
1 1 1 ln 1
u ′ − u = y ∫ − dy = − ln y = ln e y
=
y y y y
1 1
u′ − 2 u = 1
y y
′
1  1
 u  = 1 u = y + C1 u = y 2 + C1 y
 y  y

dy 1
= y 2 + C1 y dy = dx
dx y ( y + C1 )
1 1 1 
 − dy = dx
C1  y y + C1 
1 1
ln y − ln ( y + C1 ) = x + C2
C1 C1
1 y
ln = x + C2
C1 y + C1
y
ln = C1 x + C1C2
y + C1
y
= eC1 x × eC1C 2 = eC1 xC3
y + C1

16
Euler’s Equation
1 1
y′′ + Ay′ + 2 By = 0
x x
x = et or, equivalently t = ln ( x )
( )
Y ( t ) = y et
dY dt 1
y′( x ) = = Y ′( t )
dt dx x
Y ′( t ) = xy′( x )
d d 1 
y′′( x ) = y′( x ) =  Y ′( t ) 
dx dx  x 
1 1 d
= − 2 Y ′( t ) + Y ′( t )
x x dx
1 1 dY ′ dt
= − 2 Y ′( t ) +
x x dt dx
1 1 1
= − 2 Y ′( t ) + Y ′′( t )
x x x
1
= 2 ( Y ′′( t ) − Y ′( t ) )
x
x y′′( x ) = Y ′′( t ) − Y ′( t )
2

x 2 y′′( x ) + Axy′( x ) + By ( x ) = 0
Y ′′( t ) − Y ′( t ) + AY ′( t ) + BY ( t ) = 0
Y ′′( t ) + ( A − 1)Y ′( t ) + BY ( t ) = 0

The Nonhomogeneous Equation y′′ + p ( x ) y′ + q( x ) y = f ( x )

The Method of Variation of Parameters
y p ( x ) = u ( x ) y1 ( x ) + v( x ) y2 ( x )

y′p = uy1′ + vy′2 + u′y1 + v′y2

u′y1 + v′y2 = 0
y′p = uy1′ + vy′2
y′p′ = u′y1′ + v′y′2 + uy1′′ + vy′2′
u′y1′ + v′y′2 + uy1′′ + vy′2′ + p ( x )( uy1′ + vy2′ ) + q( x )( uy1 + vy2 ) = f ( x )
u[ y1′′ + p( x ) y1′ + q( x ) y1 ] + v[ y2′′ + p ( x ) y2′ + q( x ) y2 ] + u′y1′ + v′y′2 = f ( x )
u′y1′ + v′y′2 = f ( x )
y2 ( x ) f ( x ) y1 ( x ) f ( x )
u′( x ) = − and v′( x ) = −
W ( x) W ( x)

The Method of Undetermined Coefficients

17
 List of function to try in the initial stage of formulating y p

f ( x) Initial Guess for y p

P( x ) Q( x )
ceax de ax
α cos( bx ) or β sin ( bx ) c cos( bx ) + d sin ( bx )
P( x ) e ax Q( x ) e ax
P( x ) cos( bx ) or P( x ) sin ( bx ) Q( x ) cos( bx ) + R( x ) sin ( bx )
P( x ) e ax cos( bx ) or P( x ) e ax sin ( bx ) Q( x ) e ax cos( bx ) + R( x ) e ax sin ( bx )

18
CHAPTER 3
The Laplace Transform
DEFINITION 3.1 Laplace Transform
The Laplace transform L[ f ] of f is a function defined by
L[ f ]( s ) = ∫ e − st f ( t ) dt ,
∞

for all s such that this integral converges.

Ex: L[1] = ∫ 0 e (1) dt

∞
− st

1 ∞ − st
e d ( − st )
− s ∫0
=

1
= − e − st 0
s
∞
]
1
[
= lim − e − sk − e 0
k →∞ s
]
1 1 
= lim −  sk − 1
k →∞ s e 
1
=
s
THEOREM 3.7 First Shifting Theorem, or Shifting in the s variable
[ ]
L[ f ]( s ) = F ( s ) ⇒ L e at f ( t ) ( s ) = F ( s − a )
Proof

[ ] ∞
L e at f ( t ) ( s ) = ∫ e at e − st f ( s ) ds
0

= ∫ e − ( s − a ) t f ( t ) dt
∞

= F ( s − a)

THROREM 3.13
Let L[ f ]( s ) = F ( s ) for s > b and suppose that F is differentiable. Then
L[ t f ( t ) ]( s ) = − F ′( s ) for s > b .
Proof

19
 − F ′( s ) =
d ∞ − st
∫
ds 0
e f ( t ) dt = ∫
∞ d

0 ds
e − st f ( t ) dt ( )
= ∫ − te − st f ( t ) dt = ∫ e − st [ − t f ( t ) ] dt
∞ ∞

0 0

= L[ − t f ( t ) ]( s )

F ′′( s ) =
d2
ds 2 ∫
∞

0
e − st f ( t ) dt = ∫
∞

0 ds 2
(
d 2 − st
)
e f ( t ) dt
∞ ∞
=∫
0
( − t )( − t ) e − st f ( t ) dt = ∫ 0 ( − t ) 2 e − st f ( t ) dt
∞
⇒ ( − 1) F ′′( s ) = ∫ t 2e − st f ( t ) dt
2
0

[ ]
= L t 2 f (t ) ( s)

COROLLARY 3.1
Let L[ f ]( s ) = F ( s ) for s > b and let n be a positive integer. Suppose F is n times
differentiable. Then, for s > b ,

[ ]
n
n d
L t n f ( t ) ( s ) = ( − 1) F ( s)
ds n

Ex:
1
(1) L[1] =
s

(2) [ ]
L e at 1 =
1
s − a (Theorem 3.7)

L[ cos at + i sin at ] = L eiat = [ ]

1
=
s + ai
s − ai ( s − ai )( s + ai )
s a
= 2 +i 2
s +a 2
s + a2
s a
(3) ⇒ L[ cos at ] = s 2 + a 2 L[ i sin at ] = i s 2 + a 2

(4) [ ]
L t n ×1 = ( − 1)
n dn 1
n
n!
= n+1
ds s s

(5)
 e at − e − at  1
L[ sinh at ] = L 
2 
1 1
= L e at − L e − at =  ( [ ] [ ])
−
1 

2 s −a s + a
  2

20
 s + ai 1 s+a−s+a a
= = = 2
( s − ai )( s + ai ) 2 s − a
2 2
s − a2

 e at + e − at  1 1 1  s
L[ cosh at ] = L   ==  + = 2
 2  2  s − a s + a  s − a2

(6) [
L te at sinh bt ]
d  b  b( − 2 s ) 2bs
L[ t sinh bt ] = ( − 1)  2 2 
=− 2 = 2
ds  s + b  s +b 2
(2
) (
s + b2 ) 2

2b( s − a )
[
L e at t sinh bt = ]
[( s − a ) 2
+ b2 ] 2

1
L  ∫ g ( r ) dr  = L[ g ]
t
1.  0  s
1 
L−1  L[ g ]  = ∫ g ( r ) dr
t

s  0

−1  1 
Ex: L  s 2 + 1 = sin t
 
 1 
 = ∫ 0 sin rdr = [ − cos r ] 0 = − cos t − ( − cos 0) = 1 − cos t
t t
L−1  2
(
s s +1  )
lim F ( s ) = 0
2. s →∞

L[ f ′( t ) ] = sF [ s ] − f ( 0)
L[ f ′′( t ) ] = sL[ f ′] − f ′( 0 ) = s[ sF ( s ) − f ( 0) ] − f ′( 0 ) = s 2 F ( s ) − sf ( 0) − f ′( 0 )
L[ f ′′′( t ) ] = sL[ f ′′] − f ′′( 0) = s 3 F ( s ) − s 2 f ( 0 ) − sf ′( 0) − f ′′( 0 )
lim f ( t ) = lim sf ( s )
3. Initial-value theorem t →0 s →∞

lim f ( t ) = lim sf ( s )
Final-value theorem t →∞ s →0

4. If f ( 0 ) = 0 or lim sF ( s ) = 0 ⇒ L−1 [ sF ( s ) ] = f ′( t )
s →∞

If f ( 0 ) ≠ 0 ⇒ L−1 [ sF ( s ) ] = f ′( t ) + L−1 f ( 0 )
−1  as  −1  as 
Ex: L  s 2 − a 2  ⇒ L  s 2 − a 2  = sinh at = f ( t )
   
as
f ( 0) = lim =0
s →∞ s − a 22

 as  d
∴ L−1  2 = f ( t ) = a cosh at
 s − a  dt
2

21
DEFINITION 3.4 Heaviside Function
The Heaviside function H is defined by
0 if t < 0
H (t) = 
1 if t ≥ 0

DEFINITION 3.5 Pulse

A pulse is a function of the form
H ( t − a ) − H ( t − b),
in which a < b .

0 if t < 1
[ H ( t − 1) − H ( t − 2) ] et = et if 1 ≤ t < 2 .
0 if t ≥ 2


THROREM 3.8 Second Shifting Theorem, or Shifting in the t Variable

Let L[ f ]( s ) = F ( s ) for s > b Then
L[ H ( t − a ) f ( t − a ) ]( s ) = e − as F ( s )
for s > b.

22
 DEFINITION 3.6 Convolution
If f and g are defined on [ 0, ∞ ) , then the convolution f ∗ g of f with g is the function
defined by

( f ∗ g )( t ) = ∫ 0 f ( t − τ ) g (τ ) dτ
t

= ∫ f (τ ) g ( t − τ ) dτ
t

for t > 0 .

THROREM 3.9 Convolution

If f ∗ g is defined, then
L[ f ∗ g ]( s ) = L[ f ] L[ g ]

Unit Impulses and the Dirac Delta Function

1
δε ( t ) = [ H ( t ) − H ( t − ε ) ]
ε

23
 δ ( t ) = lim δ ε ( t )
ε →0+

1
δε ( t − a ) = [ H (t − a) − H (t − a − ε )]
ε

L[δ ε ( t − a ) ] =
1  1 − as 1 − ( a + ε ) s  e − as 1 − e −εs
e − e
( )
ε  s s  = εs

L[δ ( t ) ] = lim
e − as 1 − e −εs(= e − as
)
ε →0+ εs
If a=0 L[δ ( t ) ] = 1
Periodic function
1
f ( t + T ) = f ( t ) ⇒ L[ f ]( s ) = e − st f ( t ) dt
T

1 − e − sT ∫ 0

L[ f ( t ) ] = ∫ e − st f ( t ) dt = ∫ e − st f ( t ) dt + ∫
∞
e − st f ( t ) dt + ∫ e − st f ( t ) dt + 
T 2T 3T

0 0 T 2T

↑ t = u +T ↑ t = u + 2T

= ∫ e − su f ( u ) du + ∫ e − s ( u +T ) f ( u + T ) du + ∫ e − s ( u + 2T ) f ( u + 2T ) du + 
T T T

0 0 0

= ∫ e − su f ( u ) du + e − sT ∫ e − su f ( u ) du + e − 2 sT ∫ e − su f ( u ) du + 
T T T

0 0 0

( )
= 1 + e − sT + e − 2 sT +  ∫ e − su f ( u ) du
T

1
e − st f ( t ) dt
T
=
1 − e − sT ∫ 0

w  −s 
π π π π
1 w ( iw − s ) t (
Im ∫ w e − st eiwt dt = Im ∫ w e( iw − s ) t dt = Im ∫ e d iw − s ) = 2 
1 + e w

0 0 iw − s 0 w +s 
2


24
Ex: (1) Compute
 3s − 1 
L−1  2
 s − 6s + 2 
.

3s − 1 3s − 1
=
s − 6s + 2 ( s − 3) 2 − 7
2

3( s − 3) 8
= + = G ( s − 3) + K ( s − 3)
( s − 3) − 7 ( s − 3) 2 − 7
2

if we choose
3s 8
G= and K=
s −7
2
s −72

 3s − 1 
L−1  2  = L−1[ G ( s − 3) ] + L−1 [ K ( s − 3) ]
 s − 6s + 2 
= e3t L−1[ G ( s ) ] + e3t L−1 [ K ( s ) ]
 s   1 
= 3e3t L−1  2  + 8e3t L−1  2
s − 7  s − 7 
( )
= 3e3t cosh 7t +
8 3t
7
e sinh 7t ( )

Ex: (2) Compute L[ g ] , where g ( t ) = 0 for 0 ≤ t < 2 and g ( t ) = t 2 + 1 for t ≥ 2

t 2 + 1 = ( t − 2 + 2 ) + 1 = ( t − 2) − 4( t − 2 ) + 5
2 2

Then ( )
g ( t ) = t 2 + 1 H ( t − 2)

= ( t − 2 ) H ( t − 2 ) + 4( t − 2) H ( t − 2 ) + 5 H ( t − 2 )
2

Now we can apply the second shifting throrem:

[ ]
L[ g ] = L ( t − 2) H ( t − 2 ) + 4 L[ ( t − 2 ) H ( t − 2 ) ] + 5 L[ H ( t − 2 ) ]
2

[ ]
= e − 2 s L t 2 + 4e − 2 s L[ t ] + 5e − 2 s L[1]
 2 4 5
= e− 2s  3 + 2 + 
s s s

25
 Heaviside’s Formulas
Let P and Q be polynomials, and let f = L−1 [ P Q ] .

1. If Q( s ) has a factor s − a , but not ( s − a ) , then f ( t ) contains the corresponding

2

term Z ( a ) e at , where Z ( s ) = P ( s ) Q′( s ) = P ( s )( s − a ) Q( s )

2. If k ≥ 2 , and Q ( s ) has a factor ( s − a ) , but not ( s − a ) , then f ( t ) contains the

k k +1

corresponding term
 Z ( k −1) ( a ) Z ( k − 2 ) ( a ) Z ( k − 3) ( a ) t 2 Z ′( a ) t k − 2 t k −1  at
+
 ( k − 1) ! ( k − 2 ) ! t + +  + + Z ( a ) e
 ( k − 3) ! 2! 1 ! ( k − 2) ( k − 1) !
in which
P ( s )( s − a )
k
Z ( s) =
Q( s )

3. If Q( s ) contains the irreducible quadratic factor ( s − a ) + b , but not the square of

2 2

this factor, then f ( t ) contains the term

1 at
e [ β cos( bt ) + α sin ( bt ) ] ,
b
in which α and β are defined by the equations
Z ( s) =
P( s ) ( s − a ) + b 2
2
[ ]
Q( s )
and
Z ( a + ib ) = α + iβ

Ex: (3) Compute

 s 3 − 5s 2 + 2 s − 5 
L−1  2
(
 s + 1 ( s − 2)  )2 

(s 3
)
− 5s 2 + 2 s − 5 ( s − 2 )
2
=
(
 s 3 − 5s 2 + 2 s − 5 
=
)
( 8 − 20 + 4 − 5) = − 13
(a) ( 2
)
s + 1 ( s − 2)
2 
 s +1
2
( 
 s =2) 5 5

26
 Z ( k −1) ( a ) Z ′( a ) d s 3 − 5s 2 + 2 s − 5
= = Z ′( a ) =
( k − 1) ! ( 2 − 1) ! ds s2 + 1

=
( )( ) (
 s 2 + 1 3s 2 − 10s + 2 − s 3 − 5s 2 + 2s − 5 ( 2s ) 

)
 s2 + 1(2
)  s = 2
5( 3 × 4 − 10 × 2 + 2 ) − ( − 13)( 4) 5 × ( − 6 ) + 52 22
= = =
52 25 25

Z=
(s 3
− 5s 2 + 2 s − 5 s 2 + 1
=
)( ) (
s 3 − 5s 2 + 2 s − 5 ) s 2 + 1 = ( s − 0 ) + 12
2

( )
s 2 + 1 ( s − 2)
2
( s − 2) 2 a = 0 b =1

i 3 − 5( i ) + 2i − 5 − i + 5 + 2i − 5
2
i
Z ( 0 + i) = = 2 =
( i − 2) 2
i − 4i + 4 3 − 4i

i ( 3 + 4i ) − 4 + 3i −4 3
= = α= β=
( 3 − 4i )( 3 + 4i ) 25 25 25
1 at 3 4 
e [ β cos( bt ) + α sin ( bt ) ] = 1e0t  cos( t ) − sin ( t ) 
b  25 25 
3 4
= cos( t ) − sin ( t )
25 25

 s 3 − 5s 2 + 2 s − 5  22 2t 13 2t 3 4
L−1  2 = e − te + cos( t ) − sin ( t )
( )2 
 s + 1 ( s − 2)  25 5 25 25

s 3 − 5s 2 + 2 s − 5 A B Cs + D
(b) s 2 + 1 ( s − 2) 2 = s − 2 + ( s − 2) 2 + s 2 + 1
( )
 s 3 − 5s 2 + 2 s − 5 2  s 3 − 5s 2 + 2 s − 5  13
B= 2 ( s − 2)  =  =−
(
 s + 1 ( s − 2)
2
)  s =2  s +1
2 
 s=2 5

− 13
−5 A 5+D
s=0 = + (1)
4 −2 4 1

− 13
−7 A 5 +C+D
s =1 = + (2)
2 −1 1 2

− 13
− 13 A 5 + −C + D
s = −1 = + (3)
18 −3 9 2

27
 13 − 12
− 2 A + 4 D = −5 + =
5 5
26 − 9
− 2 A + C + D = −7 + =
5 5
− 13 × 6 −513 × 6 − 78 + 13×512
− 2 A − 3C + 3D = − =
18 9 18
22 −4 3
⇒ A= D= C=
25 25 25
 s 3 − 5s 2 + 2 s − 5  −1  25 
22
−1 
−13
 −1  25 s 
3
−1  25 
−4
L−1  2 = L + L 5
+ L + L
( )
 s + 1 ( s − 2) 
2  s − 2
 
 2
 ( s − 2) 
 s 2 + 1
 
 s 2 + 1
 
22 13 3 4
= e 2t − te 2t + cos( t ) − sin ( t )
25 5 25 25

Ex: (4) Compute

 s2 
L−1  
 ( s − 3) s + 5 
2
( )
s2
= sF ( s ) F ( s) =
s F ( 0 ) = sF ( s ) = 0
(
( s − 3) s 2 + 5 ) ( s − 3) s 2 + 5 ( ) s →∞

0
( )
F ( t ) = ∫ e3( t +τ ) cos 5τ dτ
t
{
= Re e3t ∫ e − 3τ ei
t

0
5τ
dτ }
 e 3t
= Re  ∫
t
(

e ( − 3 + 5i )τ d − 3 + 5i τ  )
 − 3 + 5i 0


= Re 
14
e e [
 − 3 − 5i 3t ( − 3+ 5i )τ t 
0 ]
 
 − 3 − 5i 3t − 3t
= Re  [ (
e e cos 5t + i sin 5t − 1 

) ]
 14 
3 5 3
=− cos 5t + sin 5t + e3t
14 14 14

−1  s2  9 3t 3 5 5
L−1 [ sF ( s ) ] = f ′ ⇒ L   = 14 e + 14 sin 5t + 14 cos 5t
 ( s − 3) s + 5
2
( ) 

Ex: (5) Determine f such that

f ( t ) = 2t 2 + ∫ f ( t − τ ) e −τ dτ
t

f ( t ) = 2t 2 + f ( t ) ∗ e −t
Taking the Laplace transform of this equation yields

28
 4 1
F ( s) = + F ( s)
s 3
s +1
Then
4 4
F ( s) = +
s3 s 4
and from this we easily invert to obtain
2
f ( t ) = 2t 2 + t 3
3

Ex: (6) Suppose we want to solve

y′′ + 2ty′ − 4 y = 1 ; y ( 0 ) = y′( 0) = 0
Apply the Laplace transform to get
1
s 2Y ( s ) − sy ( 0 ) − y′( 0 ) + 2 L[ ty′] ( s ) − 4Y ( s ) = .
s
Now y ( 0 ) = y′( 0 ) = 0 and
d
L[ ty′]( s ) = − [ L[ y′]( s ) ] = − d [ sY ( s ) − y( 0) ] = −Y ( s ) − sY ′( s )
ds ds
We therefore have
1
s 2Y ( s ) − 2Y ( s ) − 2sY ′( s ) − 4Y ( s ) =
s
3 s 1
Or Y ′ +  − Y = − 2
 s 2 2s
3 s  1 2
∫  s − 2  ds = 2 ln( s ) − 4 s
The exponential of this function, or 2 ,
s 3e − s 4

(s e
3 −s2 4
)
′ 1
Y = − se − s
2
2
4

Then 2 2
s 3e − s 4Y = e − s 4
+C

1 C s2
so Y ( s) = + e 4

s3 s3
However, in order to have lim
s→∞ Y ( s ) = 0 , we must choose C=0.

1 2
Then Y ( s ) = 1 s so y ( t ) = 2 t
3

29