Professional Documents
Culture Documents
ME233 Sp16 CT1 Random Vector Processes
ME233 Sp16 CT1 Random Vector Processes
ME233 Sp16 CT1 Random Vector Processes
Random processes
Random Process
A random processes is a continuous function
of time
x1(t)
0.1
Example
0
-0.1
0 0.5 1 1.5 2 2.5 3
0.1
x2(t)
Ensemble
0
-0.1
0 0.5 1 1.5 2 2.5 3
0.1
x3(t)
-0.1
0 0.5 1 1.5 2 2.5 3
0.1
x4(t)
-0.1
0 0.5 1 1.5 2 2.5 3
t
t
4
x1(t)
0.1
Example
0
-0.1
0 0.5 1 1.5 2 2.5 3
0.1
sample function
x2(t)
Ensemble
0
process realization
-0.1
0 0.5 1 1.5 2 2.5 3
0.1
x3(t)
-0.1
0 0.5 1 1.5 2 2.5 3
0.1
x4(t)
-0.1
0 0.5 1 1.5 2 2.5 3
t
t
5
Random process
Let be a random process
Auto-covariance function:
7
Auto-covariance function
Define:
8
Ergodicity
A Strict Sense Stationary random process
with probability 1
(almost surely)
10
0.1
x1(t)
-0.1
0 0.5 1 1.5 2 2.5 3
Ensemble
0.1
x2(t)
-0.1
0 0.5 1 1.5 2 2.5 3
0.1
x3(t)
-0.1
0 0.5 1 1.5 2 2.5 3
0.1
x4(t)
-0.1
0 0.5 1 1.5 2 2.5 3
t
t
11
Notice that:
14
Cross-covariance function
Let and
be two WSS random vector processes
Cross-covariance function
16
White noise
A WSS random process is white if:
White noise
The power spectral density function for white noise
is:
Proof:
1
20
White noise
WW ()
WW (w)
Infinite bandwidth
0 w
21
where
Let
is also WSS
24
G(t)
Then:
G()
26
G(s)
G(w)
27
G(s)
28
If
Then:
29
Proof:
Then:
30
G(t)
Then:
G()
31
G(s)
32
If
Then:
34
then
and
35
with
41
Also,
where:
42
Also,
where:
43
and A Hurwitz,
44
Satisfies:
45
Approximate
Numerical Integration
The state equation
where
55
Thus,
where
56
t t
Thus,
59
1) Lets calculate
using
61
using
62
ΔT
1
ΔT
0
64
2) Thus
and
67
Notice that:
where,
68