Global Positioning System Reference
In-Depth Information
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
It follows from the law of variance propagation (4.34) and (4.53) that
P 1 B T M 1 M
A A T M 1 A 1 A T M 1 BP 1
=
Q v
(4.56)
The adjusted observations are
a = b +
v
= b + P 1 B T M 1 A A T M 1 A 1 A T M 1
(4.57)
P 1 B T M 1 w
Be cause
P 1 B T M 1 A A T M 1 A 1 A T M 1 B
a
P 1 B T M 1 B
b =
I
+
(4.58)
[10
it follows that
Lin
* 1 ——
Lon
PgE
Q
=
Q
b
Q v
(4.59)
a
where the inverse of P has been replaced by Q
according to (4.4)
b
4. 4.4 A Posteriori Variance of Unit Weight
The minimum of v T Pv follows from (4.49), (4.50), and (4.52) as
[10
w T M 1
M 1 A A T M 1 A 1 A T M 1 ] w
v T Pv
=
(4.60)
Th e expected value of this random variable is
E v T Pv =
E Tr v T Pv
E Tr w T M 1
M 1 A A T M 1 A 1 A T M 1 w
=
E Tr M 1
M 1 A A T M 1 A 1 A T M 1 ww T
(4.61)
=
Tr M 1
M 1 A A T M 1 A 1 A T M 1 E ww T
=
Th e trace (Tr) of a matrix equals the sum of its diagonal elements. In the first part
of (4.61), the property that the trace of a 1
1 matrix equals the matrix element
its elf is used. Next, the matrix products are switched, leaving the trace invariant. In
th e last part of the equation, the expectation operator and the trace are switched. The
expected value E( ww T ) can be readily computed. Per definition, the expected value
of the residuals
×
E( v )
=
o
(4.62)
 
Search WWH ::




Custom Search