The roots of the equation z2 + 4 z + 1 = 0 are −2 ±√3 of which
only −2+√3 is inside the unit circle so
=
2
i
2πiRes
⎛ ⎝
1
z2 + 4 z + 1
;−2 + √3
⎞ ⎠
=
4π
1
2 z + 4
|z = −2 + √3
=
2π
√3
.
(2)
Note that we used the short-cut to find a residue at a first order
pole. Example
⌠ ⌡
2π
0
cos6 θdθ.
(Solution)
⌠ ⌡
2π
0
cos6 θdθ
=
⌠ (⎜) ⌡
|z|=1
⎛ ⎝
1
2
(z +
1
z
)
⎞ ⎠
6
dz
iz
=
1
64i
⌠ (⎜) ⌡
|z|=1
(1 + z2)6
z7
dz
=
1
64i
⌠ (⎜) ⌡
|z|=1
z12 + 6 z10 + 15 z8 + 20 z6 + 15 z4 + 6 z2 + 1
z7
dz
=
1
64i
⎛ ⎝
…+
⌠ (⎜) ⌡
|z|=1
20
dz
z
+ …
⎞ ⎠
=
1
64i
20
⌠ (⎜) ⌡
|z|=1
dz
z
=
1
64i
×20 ×2πi
=
5
8
π.
(3)
Example 1:
⌠ ⌡
∞
−∞
dx
1 + x2
.
The integrand satisfies the above two conditions (the difference in
the polynomial order is 2 and z = ±i are not on the real axis. Note
that only z = i is in the upper half plane.) so
⌠ ⌡
∞
−∞
dx
1 + x2
=
2πiRes(
1
1 + z2
;i)
=
2πi
1
2 z
|z=i
=
π.
(4)
Of course this result can be also obtained by directly integrating
1/(1 + x2) (=arctanx). Example 2:
⌠ ⌡
∞
−∞
dx
1 + x4
dx.
(5)
The difference in degree of P(x)/Q(x) is 4 and all the zeros
to 1 + z4 = 0 are off the real axis. The zeros to z4 + 1 = 0 are
z = eπi/4 + n πi/2 n=0,1,2,3,
of which z's for n = 0 and n = 1 are in the upper half plane.
⌠ ⌡
∞
−∞
dx
1 + x4
=
2πi
∑ upperhalfplane
Res(
1
z4+1
;ai)
=
2πi
⎛ ⎝
Res(
1
z4 + 1
; eπi/4) +Res(
1
z4 + 1
; e3πi/4)
⎞ ⎠
=
2πi
⎛ ⎝
1
4z3
|z=eπi/4 +
1
4z3
|z=e3πi/4
⎞ ⎠
=
2πi
4
⎛ ⎝
1
e3πi/4
+
1
eπi/4
⎞ ⎠
=
πi
2
(e−3πi/4 + e−π/4 i)
=
πi
2
( −√2i )
=
π
√2
.
(6)
Example 3:
⌠ ⌡
∞
−∞
x2dx
1 + x4
=
2πi
∑ upperhalfplane
Res(
z2
z4 + 1
;ai)
=
2πi
⎛ ⎝
z2
4 z3
|z=exp(πi/4) +
z2
4 z3
|z=exp(3πi/4)
⎞ ⎠
=
πi
2
⎛ ⎝
1
z
|z=exp(πi/4) +
1
z
|z=exp(3πi/4)
⎞ ⎠
=
πi
2
(e−πi/4 + e−3πi/4)
=
πi
2
( −√2 i)
=
π
√2
.
(7)
File translated from
TEX
by
TTH,
version 4.03. On 18 Jan 2021, 22:02.