Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Complex Numbers and Exponentials: Definition and Basic Operations

Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

Complex Numbers and Exponentials

Definition and Basic Operations

A complex number is nothing more than a point in the xy–plane. The sum and product of two complex
numbers (x1 , y1 ) and (x2 , y2 ) is defined by

(x1 , y1 ) + (x2 , y2 ) = (x1 + x2 , y1 + y2 )


(x1 , y1 ) (x2 , y2 ) = (x1 x2 − y1 y2 , x1 y2 + x2 y1 )

respectively. It is conventional to use the notation x + iy (or in electrical engineering country x + jy) to stand
for the complex number (x, y). In other words, it is conventional to write x in place of (x, 0) and i in place of
(0, 1). In this notation, the sum and product of two complex numbers z1 = x1 + iy1 and z2 = x2 + iy2 is given
by
z1 + z2 = (x1 + x2 ) + i(y1 + y2 )
z1 z2 = x1 x2 − y1 y2 + i(x1 y2 + x2 y1 )
The complex number i has the special property

i2 = (0 + 1i)(0 + 1i) = (0 × 0 − 1 × 1) + i(0 × 1 + 1 × 0) = −1

For example, if z = 1 + 2i and w = 3 + 4i, then

z + w = (1 + 2i) + (3 + 4i) = 4 + 6i
zw = (1 + 2i)(3 + 4i) = 3 + 4i + 6i + 8i2 = 3 + 4i + 6i − 8 = −5 + 10i

Addition and multiplication of complex numbers obey the familiar algebraic rules

z1 + z2 = z2 + z1 z1 z2 = z2 z1
z1 + (z2 + z3 ) = (z1 + z2 ) + z3 z1 (z2 z3 ) = (z1 z2 )z3
0 + z1 = z1 1z1 = z1
z1 (z2 + z3 ) = z1 z2 + z1 z3 (z1 + z2 )z3 = z1 z3 + z2 z3

The negative of any complex number z = x + iy is defined by −z = −x + (−y)i, and obeys z + (−z) = 0.

Other Operations

The complex conjugate of z is denoted z̄ and is defined to be z̄ = x − iy . That is, to take the complex
conjugate, one replaces every i by −i. Note that

z z̄ = (x + iy)(x − iy) = x2 − ixy + ixy + y 2 = x2 + y 2

is always a positive real number. In fact, it is the square of the distance from x + iy (recall that this is the point
(x, y) in the xy–plane) to 0 (which is the point (0, 0)). The distance from z = x + iy to 0 is denoted |z| and is
called the absolute value, or modulus, of z . It is given by
p √
|z| = x2 + y 2 = z z̄


c Joel Feldman. 2003. All rights reserved. 1
Since z1 z2 = (x1 + iy1 )(x2 + iy2 ) = (x1 x2 − y1 y2 ) + i(x1 y2 + x2 y1 ),
p
|z1 z2 | =(x1 x2 − y1 y2 )2 + (x1 y2 + x2 y1 )2
q
= x21 x22 − 2x1 x2 y1 y2 + y12 y22 + x21 y22 + 2x1 y2 x2 y1 + x22 y12
q q
= x21 x22 + y12 y22 + x21 y22 + x22 y12 = (x21 + y12 )(x22 + y22 )
= |z1 ||z2 |

for all complex numbers z1 , z2 .


Since |z|2 = z z̄, we have z |z|z̄ 2 = 1 for all complex numbers z 6= 0 . This says that the multiplicative


inverse, denoted z −1 or z1 , of any nonzero complex number z = x + iy is

x−iy y
z −1 = z̄
|z|2 = x2 +y 2 = x
x2 +y 2 − x2 +y 2 i

It is easy to divide a complex number by a real number. For example

11+2i 11 2
25 = 25 + 25 i

In general, there is a trick for rewriting any ratio of complex numbers as a ratio with a real denominator. For
1+2i
example, suppose that we want to find 3+4i . The trick is to multiply by 1 = 3−4i
3−4i . The number 3 − 4i is the
complex conjugate of 3 + 4i. Since (3 + 4i)(3 − 4i) = 9 − 12i + 12i + 16 = 25

1+2i 1+2i 3−4i (1+2i)(3−4i) 11+2i 11 2


3+4i = 3+4i 3−4i = 25 = 25 = 25 + 25 i

The notations Re z and Im z stand for the real and imaginary parts of the complex number z, respec-
tively. If z = x + iy (with x and y real) they are defined by

Re z = x Im z = y

Note that both Re z and Im z are real numbers. Just subbing in z̄ = x − iy gives

Re z = 21 (z + z̄) Im z = 1
2i (z − z̄)

The Complex Exponential

Definition and Basic Properties. For any complex number z = x + iy the exponential ez , is defined by

ex+iy = ex cos y + iex sin y

In particular, eiy = cos y + i sin y. This definition is not as mysterious as it looks. We could also define eiy by
P∞ n
the subbing x by iy in the Taylor series expansion ex = n=0 xn! .

(iy)2 (iy)3 (iy)4 (iy)5 (iy)6


eiy = 1 + iy + 2! + 3! + 4! + 5! + 6! + ···

The even terms in this expansion are

(iy)2 (iy)4 (iy)6 y2 y4 y6


1+ 2! + 4! + 6! + ··· = 1 − 2! + 4! − 6! + · · · = cos y


c Joel Feldman. 2003. All rights reserved. 2
and the odd terms in this expansion are
 
(iy)3 (iy)5 y3 y5
iy + 3! + 5! + ··· = i y − 3! + 5! + ··· = i sin y

For any two complex numbers z1 and z2

ez1 ez2 = ex1 (cos y1 + i sin y1 )ex2 (cos y2 + i sin y2 )


= ex1 +x2 (cos y1 + i sin y1 )(cos y2 + i sin y2 )
= ex1 +x2 {(cos y1 cos y2 − sin y1 sin y2 ) + i(cos y1 sin y2 + cos y2 sin y1 )}
= ex1 +x2 {cos(y1 + y2 ) + i sin(y1 + y2 )}
= e(x1 +x2 )+i(y1 +y2 )
= ez1 +z2

so that the familiar multiplication formula also applies to complex exponentials. For any complex number
c = α + iβ and real number t
ect = eαt+iβt = eαt [cos(βt) + i sin(βt)]

so that the derivative with respect to t


d ct
dt e = αeαt [cos(βt) + i sin(βt)] + eαt [−β sin(βt) + iβ cos(βt)]
= (α + iβ)eαt [cos(βt) + i sin(βt)]
= cect

is also the familiar one.

Relationship with sin and cos. When θ is a real number

eiθ = cos θ + i sin θ


e−iθ = cos θ − i sin θ = eiθ

are complex numbers of modulus one. Solving for cos θ and sin θ (by adding and subtracting the two equations)

cos θ = 12 (eiθ + e−iθ ) = Re eiθ



sin θ = 1
2i (e − e−iθ ) = Im eiθ

These formulae make it easy derive trig identities. For example

cos θ cos φ = 14 (eiθ + e−iθ )(eiφ + e−iφ )


= 14 (ei(θ+φ) + ei(θ−φ) + ei(−θ+φ) + e−i(θ+φ) )
= 14 (ei(θ+φ) + e−i(θ+φ) + ei(θ−φ) + ei(−θ+φ) )
= 12 cos(θ + φ) + cos(θ − φ)


and, using (a + b)3 = a3 + 3a2 b + 3ab2 + b3 ,


3
sin3 θ = − 8i
1
eiθ − e−iθ
1
ei3θ − 3eiθ + 3e−iθ − e−i3θ

= − 8i
1
= 43 2i eiθ − e−iθ − 41 2i
1
ei3θ − e−i3θ
 

3 1
= 4 sin θ − 4 sin(3θ)


c Joel Feldman. 2003. All rights reserved. 3
Polar Coordinates. Let z = x + iy be any complex number. Writing (x, y) in polar coordinates in the usual
way gives x = r cos θ, y = r sin θ and

y
x + iy = reiθ

x + iy = r cos θ + ir sin θ = reiθ r


θ
x

In particular

y
i=(0,1)
1 = ei0 = e2πi = e2kπi for k = 0, ±1, ±2, · · ·
π
(−1,0)=−1 π 2 1=(1,0) −1 = eiπ = e3πi = e (1+2k)πi
for k = 0, ±1, ±2, · · ·
5 1
x i = eiπ/2 = e 2 πi = e( 2 +2k)πi for k = 0, ±1, ±2, · · ·
− π2 3 1
−i = e−iπ/2 = e 2 πi = e(− 2 +2k)πi for k = 0, ±1, ±2, · · ·
−i=(0,−1)

y
The polar coordinate θ = tan−1 x associated with the complex number z = x + iy is also called the argument
of z.
The polar coordinate representation makes it easy to find square roots, third roots and so on. Fix any
positive integer n. The nth roots of unity are, by definition, all solutions z of

zn = 1

Writing z = reiθ
rn enθi = 1e0i

The polar coordinates (r, θ) and (r′ , θ′ ) represent the same point in the xy–plane if and only if r = r′ and
θ = θ′ + 2kπ for some integer k. So z n = 1 if and only if rn = 1, i.e. r = 1, and nθ = 2kπ for some integer k.
k
The nth roots of unity are all complex numbers e2πi n with k integer. There are precisely n distinct nth roots
k k ′ ′ ′
of unity because e2πi n = e2πi n if and only if 2π nk − 2πi kn = 2π k−k
n is an integer multiple of 2π. That is, if

and only if k − k is an integer multiple of n. The are n distinct nth roots of unity are

y
2πi 26 1
e e2πi 6
3
1 2 3 n−1
e2πi 6 =−1 2πi 60
1 , e2πi n , e2πi n , e2πi n , · · · , e2πi n 1=e
x

4 5
e2πi 6 e2πi 6


c Joel Feldman. 2003. All rights reserved. 4
Exploiting Complex Exponentials in Calculus Computations
Example 1 Z Z Z
ex cos x dx = 1
ex eix + e−ix dx = 1
e(1+i)x + e(1−i)x dx
   
2 2

1 1 (1+i)x 1 (1−i)x
 
= 2 1+i e + 1−i e +C
This form of the indefinite integral looks a little wierd because of the i’s. But it is correct and it is purely
1 (1−i)x 1 (1+i)x
real, despite the i’s, because 1−i e is the complex conjugate of 1+i e . We can convert the indefinite
1 1−i
integral into a more familar form just by subbing back in e ±ix
= cos x ± i sin x, 1+i = (1+i)(1−i) = 1−i
2 and
1 1 1+i
1−i = 1+i = 2 . Z
ex cos x dx = 12 ex 1 ix 1 −ix
 
1+i e + 1−i e +C
x 1−i
= 21 e 1+i
 
2 (cos x + i sin x) + 2 (cos x − i sin x) + C
= 21 ex cos x + 12 ex sin x + C

Example 2 Using (a + b)4 = a4 + 4a3 b + 6a2 b2 + 4ab3 + b4 ,


Z Z Z
4
cos4 x dx = 214
 ix
e + e−ix dx = 214
 4ix
e + 4e2ix + 6 + 4e−2ix + e−4ix dx


 1 4ix 4 2ix 4 −2ix 1 −4ix


= 214 4i

e + 2i e + 6x + −2i e + −4i e +C
= 214 21 2i
1
(e4ix − e−4ix ) + 2i
4
(e2ix − e−2ix ) + 6x + C
 

= 214 21 sin 4x + 4 sin 2x + 6x + C


 

1 1
= 32 sin 4x + 4 sin 2x + 83 x + C

Example 3 We shall now guess a solution to the differential equation

y ′′ + 2y ′ + 3y = cos t (1)

Equations like this arise, for example, in the study of the RLC circuit. We shall simplify the computation by
exploiting that cos t = Re eit . First, we shall guess a function Y (t) obeying

Y ′′ + 2Y ′ + 3Y = eit (2)

Then, taking complex conjugates,


Ȳ ′′ + 2Ȳ ′ + 3Ȳ = e−it (2̄)
1 1
and, adding 2 (2) and 2 (2̄) together will give

(Re Y )′′ + 2(Re Y )′ + 3(Re Y ) = Re eit = cos t

which shows that Re Y (t) is a solution to (1). Let’s try Y (t) = Aeit . This is a solution of (2) if and only if
d2 it
+ 2 ddt Aeit + 3Aeit = eit
 
dt2 Ae
⇐⇒ (2 + 2i)Aeit = eit
1
⇐⇒ A= 2+2i
it
e
So we have found a solution to (2) and Re 2+2i is a solution to (1). To simplify this, write 2 + 2i in polar
coordinates. So
√ π eit it π
eit
2 + 2i = 2 2ei 4 ⇒ 2+2i = √e i π4 = 2√1
2
ei(t− 4 ) ⇒ Re 2+2i 1
= 2√ 2
cos(t − π4 )
2 2e


c Joel Feldman. 2003. All rights reserved. 5
R√
Example 4 In this example, we shall find x2 − 1 dx.
R√
First, here is some motivation for the substitution that I shall use. To integrate√ 1 − x2 dx, we
√ √
substitute x = cos t, since it is easy to take the square root in 1 − x2 = 1 − cos2 t = sin2 t. Now that
we know about complex numbers, we are no longer afraid of taking the p square root of negative numbers.

2

2 2
√ √ 2
Consequently, we can still substitute x = cos t into x − 1 = cos t − 1 = − sin t = −1 sin t = ±i sin t.
R√
In any real application, the domain of integration for x2 − 1 dx will only include x’s obeying x2 ≥ 1,

2
so that x − 1 is real. This looks like it causes problems for the substitution x = cos t, because we are used
to thinking that cos t only takes values between −1 and 1. But the restriction −1 ≤ cos t ≤ 1 is only valid
when t is real. Allowing t to be complex allows cos t to take all possible complex values. In fact, I claim that
as t runs over all pure imaginary values (that is t = iy with y real), cos t takes all real values bigger than +1.
To see this, set z = it. Then as t runs over all pure imaginary values, z runs over all pure real values. When
z = 0, cos t = 21 eit + e−it = 21 ez + e−z takes the value 1. As z increases, 21 ez + e−z increases (because
  
d 1 z
 1 z
−z
= 2 e − e−z > 0 for z > 0) and as z approachs infinity, so does 21 ez + e−z . Thus as z runs
 
dz 2 e + e
through the real numbers from 0 to infinity, 21 ez + e−z runs through the real numbers from 1 to infinity. The


function 21 ez + e−z is called the hyperbolic cosine of z and is denoted cosh z. Similarly, the hyperbolic sine of


z is sinh z = 21 ez − e−z . The relationship between hyperbolic and regular sine and cosine is


cos y = cosh iy i sin y = sinh iy

For every trig identity, there is a corresponding identity for sinh and cosh. Just the signs change. For example
sin2 x + cos2 x = 1, but cosh2 x − sinh2 x = 1. The identities are checked by just subbing in sinh z = 21 ez − e−z


and cosh z = 12 ez + e−z . Similarly, the derivative rules for sinh and cosh are the same as those for sin and

d
cos, up to signs. For example, while dx cos x = − sin x, ddx cosh x = sinh x.
Now the evaluation of the integral. Suppose that we want x ≥ 1. Sub in x = cosh z = 21 ez + e−z


with z ≥ 0. (If we wanted x ≤ −1, we would sub in x = − cosh z.) I’ll write everything out explicitly in terms
of exponentials. The formulae would be shorter, if I wrote everything in terms of cosh x and sinh x.
1
ez + e−z

x= 2
1
ez − e−z dz

dx = 2
−z 2
2
x2 − 1 = 1 z
− 1 = 14 e2z + 2 + e−2z − 1 = 41 e2z − 2 + e−2z = 1
ez − e−z
  
4 e +e 4
p
1 z −z

x2 − 1 = 2 e −e
−z 2
p
1 z
dz = 14 e2z − 2 + e−2z dz
 
x2 − 1 dx = 4 e −e
Z p Z
1
e2z − 2 + e−2z dz = 41 12 e2z − 2z − 21 e−2z + C
 
x2 − 1 dx = 4

1
ez + e−z for z as a

Now we have to sub back in what z is in terms of x. That is, we have to solve x = 2
function of x.

1
ez + e−z ⇐⇒ 2x = ez + e−z ⇐⇒ 2xez = e2z + 1 ⇐⇒ e2z − 2xez + 1 = 0

x= 2

Think of this as the quadratic equation Q2 − 2xQ + 1 = 0 for Q = ez . The quadratic equation Q2 − 2xQ + 1 = 0
√ √
has two solutions: Q = 12 2x ± 4x2 − 4 = x ± x2 − 1. Note that if we divide the equation e2z − 2xez + 1 = 0


by e2z we get e−2z − 2xe−z + 1 = 0, which is exactly the same quadratic equation for Q′ = e−z as we had for

Q. One of the two solutions x ± x2 − 1 is ez and the other is e−z . As we want z ≥ 0, so that ez ≥ e−z , we
√ √
have to choose ez = x + x2 − 1 and e−z = x − x2 − 1. As a check, note that
p p p
x2 − 1 x − x2 − 1 = x2 − (x2 − 1) = 1 √1
 
x+ ⇒ x+ x2 −1
=x− x2 − 1


c Joel Feldman. 2003. All rights reserved. 6
√ √ √
Subbing in ez = x + x2 − 1 and e−z = x − x2 − 1 and z = ln x + x2 − 1 ,


Z p
x2 − 1 dx = 14 21 e2z − 2z − 21 e−2z + C
 

p 2 p p 2 
= 14 21 x + x2 − 1 − 2 ln x + x2 − 1 − 12 x − x2 − 1
 
+C
√ √
= 12 x x2 − 1 − ln x + x2 − 1 + C
 

As a check, note that


 p p hp √ i
x2 −1
d 1
x x2 − 1 − ln x + x2 − 1 = 12 x2 − 1 + x √xx2 −1 − 1+x/


dx 2 x+ x2 −1
h 2 2
√ i
2 −1+x
= 12 √xx2−1
−1
+ √x
2
x −1
− √ 1
2
x√
x −1 x+ x −1 2
h i
1 √x2 −1 2
= 2 x2 −1 + √x2 −1 − √x12 −1
x

h 2 i 2
p
= 21 √2x −2
2
x −1
= √x −1 =
2
x −1
x2 − 1

as desired.
R x+2
Example 5 In this example, we shall find x2 +2x+5 dx. Using complex numbers, any polynomial can be
written as a product of linear factors. This allows us to eliminate quadratic denominators from the partial
fractions procedure. This example illustrates how.
We first have to factor the denominator x2 + 2x + 5. We can use the high school formula for the roots
√ √ √
of a quadratic equation: −2± 22 −4×5 = −2± 2 4−20 = −1 ± −4 = −1 ± 2i. Or we can complete the square
2

x2 + 2x + 5 = (x + 1)2 + 4 = (x + 1)2 − (2i)2 = [(x + 1) − 2i][(x + 1) + 2i] = [x + 1 − 2i][x + 1 + 2i]

Next we write the integrand in the form

x+2 x+2 a b
x2 +2x+5 = (x+1−2i)(x+1+2i) = x+1−2i + x+1+2i

with the constants a and b chosen so that

a b a(x+1+2i)+b(x+1−2i) x+2
x+1−2i + x+1+2i = (x+1−2i)(x+1+2i) = (x+1−2i)(x+1+2i) i.e. so that a(x + 1 + 2i) + b(x + 1 − 2i) = x + 2

This has to be true for all x. We can solve easily for a if we choose x + 1 = 2i and we can solve easily for b if
we choose x + 1 = −2i:
1+2i 1 1
x + 1 = 2i ⇒ a(2i + 2i) + b(2i − 2i) = 2i + 1 ⇒ 4i a = 1 + 2i ⇒ a= 4i = 2 − 4 i
x + 1 = −2i ⇒ a(−2i + 2i) + b(−2i − 2i) = −2i + 1 ⇒ −4i b = 1 − 2i ⇒ b= − 1−2i 1 1
4i = 2 + 4 i

1 1
since i = −i. As a check, we observe that, with a = 2 − 41 i and b = 1
2 + 41 i,
1
− 41 i (x + 1 + 2i) + 1
+ 41 i (x + 1 − 2i)
 
a(x + 1 + 2i) + b(x + 1 − 2i) = 2 2
1
− 41 i + 21 + 14 i + 2i 1
− 41 i − 1
− 41 i
 
= (x + 1) 2 2 2
= x + 1 + 2i − 12 i = x + 2


as desired. The integral is now easy,


Z Z h i
x+2 a b
x2 +2x+5 dx = x+1−2i + x+1+2i dx = a ln(x + 1 − 2i) + b ln(x + 1 + 2i) + C

= 21 − 41 i ln(x + 1 − 2i) + 21 + 41 i ln(x + 1 + 2i) + C


 


c Joel Feldman. 2003. All rights reserved. 7
though the answer looks a little wierd because of the complex numbers.
One can eliminate the complex numbers by using the fact that
p
ln(X ± iY ) = ln X 2 + Y 2 ± i tan−1 Y
X (L)

To derive (L), let ln(X ± iY ) = U ± iV , with U and V real. Then U and V are to be determined by
eU±iV = X ± iY or eU cos V ± i sin V ) = X ± iY or eU cos V = X, eU sin V = Y . Dividing the last two
Y
equations gives tan V = X and adding the squares of the last two equations together gives e2U = X 2 + Y 2 .
Applying (L) with X = x + 1 and Y = 2 gives
 p
1
− 41 i ln(x + 1 − 2i) + 1
+ 41 i ln(x + 1 + 2i) = 1
− 41 i x2 + 2x + 5 − i tan−1 x+1
2
  
2 2 2
 p
+ 12 + 41 i x2 + 2x + 5 + i tan−1 x+1
2

p
= x2 + 2x + 5 − 1
2 tan−1 2
x+1


c Joel Feldman. 2003. All rights reserved. 8

You might also like