Scribd
Upload
16 views

Advanced Topics in Diagnosis and Fault-Tolerant Control: Chapter 2.2.2 - 2.2.3

The document discusses the chi-squared probability density function (PDF) and its properties. It can be used to model the sum of squared normal random variables. The chi-squared PDF depends on the number of degrees of freedom and has specific forms whether that number is even or odd. It also generalizes the gamma and factorial functions. The Q-function probability is defined as the tail of the chi-squared distribution. Finally, a non-central chi-squared distribution is described which allows the normal variables to have non-zero means.

Uploaded by

Lidius85
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
16 views

Advanced Topics in Diagnosis and Fault-Tolerant Control: Chapter 2.2.2 - 2.2.3

The document discusses the chi-squared probability density function (PDF) and its properties. It can be used to model the sum of squared normal random variables. The chi-squared PDF depends on the number of degrees of freedom and has specific forms whether that number is even or odd. It also generalizes the gamma and factorial functions. The Q-function probability is defined as the tail of the chi-squared distribution. Finally, a non-central chi-squared distribution is described which allows the normal variables to have non-zero means.

Uploaded by

Lidius85
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 11

Advanced topics in Diagnosis and Fault-tolerant

Control
Chapter 2.2.2 - 2.2.3
Lidia Furno
1 April 2014
The
2
PDF (1)

2
is related to the standard normal distribution.
If a random variable x is Gaussian, then x
2
has a
2
probability density function (PDF) with one degree
of freedom.
The
2
PDF (2)
If x
1
, x
2
, ..., x

are independent Gaussian variables,


then
x = x
2
1
+ x
2
2
+... + x
2

(1)
has a
2
PDF with degrees of freedom.
is the number of independent Gaussian variables that we
have added up.
The Central
2
PDF (1)
The Gaussian random variables x:

are independent from one another;

are identically distributed (same pdf);

x
i
N(0, 1), i = 1, ..., .
The Central
2
PDF (2)
p(x) =

1
2

2
(

2
)
x

2
1
e

1
2
x
, x 0
0 , x < 0
(2)
where is the number of degrees of freedom
( 1) and (u) is the Gamma function.
The mean and the variance are:
E(x) = (3)
var (x) = 2 (4)
The Central
2
PDF (3)
The Gamma function
The Gamma function is dened as:
(u) =


0
t
u1
e
t
dt for any u (5)
By applying iteratively partial integration to the Gamma function:
(u) = (u 1)(u 2)...(1) = (u 1)(u 2)...1 (6)
It follows that the Gamma function is a generalization of the
factorial function:
(n) = (n 1)! (7)
where n is an integer.
Q function - even
The Q function probability for a
2
random variable is dened as:
Q
X
2 =


x
p(t)dt x > 0 (8)
In case of = 2:

1
2
e

1
2
x
, x > 0
0, x < 0
(9)
Otherwise, for even:
Q
X
2 = e

1
2
x

2
1

k=0
(
x
2
)
2
k!
(10)
Q function - odd
The Q function probability for a
2
random variable is dened as:
Q
X
2 =


x
p(t)dt x > 0 (11)
For odd:

2Q(

x), = 1
2Q(

x) +
e

1
2
x

1
2

k=1
(k 1)!(2x)
k
1
2
(2k 1)!
, 3
(12)
The Non Central
2
PDF
The Gaussian random variables x:

are independent from one another;

are identically distributed (same pdf);

x
i
N(
i
, 1), i = 1, ..., .
Two important parameters:

, degrees of freedom

i =1

2
i
, noncentrality patameter

You might also like