PROBABILITY AND QUEUEING

THEORY

By

Dr.M.PERUMALSAMY

ASSISTANT PROFESSOR

DEPARTMENT OF MATHEMATICS

ALAGAPPA CHETTIAR GOVERNMENT COLLEGE OF

ENGINEERING AND TECHNOLOGY

KARAIKUDI

[email protected]

September 8, 2020

1/ 40
 UNIT I: Random variables and Probability

distributions
Continuous Probability Dsitibutions

Uniform 0r Rectangular Distribution

A continuous random variable x is said to have a uniform dis-
tribution if its probability density function is given by
( 1
for a < x < b
f (x) = b−a
0 otherwise

Derivation: To nd Mean, Variance and Moment Generating

Function of Uniform distribution
Rb
Mean= µ1 = E (x) = a xf (x)dx
0

Rb 1
= a x dx
b−a 2/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 1 b 2 − a2
= ( )
b−a 2
b+a
=
2
Now
E (x 2 ) = a x 2 f (x)dx
Rb

1
= a x 2(
Rb
)dx
b−a
1 x3 b
= [ ]
b−a 3 a
1 b 3 − a3
= ( )
b−a 3
1 (b − a)(b2 + ab + a2 )
=
3 b−a
(b + ab + a2 )
2
E (x 2 ) =
3 3/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 ∴ Var (X ) = E (x 2 ) − [E (x)]2
(b 2 + ab + a2 ) b+a 2
= −( )
3 2
(b 2 + ab + a2 ) a2 + b 2 + 2ab 2
= −( )
3 4
(4b 2 + 4ab + 4a2 − 3a2 − 3b 2 − 6ab)
=
12
2 2
(b − 2ab + a
=
12
(b − a)2
=
12
MX (t) = E {e tx }
Rb
= a e tx f (x)dx
Rb 1
= a e tx ( )dx
b−a 4/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 1 e tx b
= [ ]
b−a t a
e bt − e at
=
t(b − a)

For Uniform distribution

( 1
for a<x <b
f (x) = b−a
0 otherwise
e bt − e at
MGF = MX (t) =
t(b − a)
b+a
Mean =
2
(b − a)2
variance =
12 5/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 Problem: 1
A random variable X has a uniform distribution over
(−3, 3), compute (i) P(X < 2), P(|X | < 2), P(|X − 2| < 2)
(ii) Find k for which P(X > k) = 1/3.
Solution:
W.K.T. the pdf of a random variable X which is distributed
uniformly in (a, b) is
( 1
for a < x < b
f (x) = b−a
0 otherwise
(
1
for − 3 < x < 3
f (x) = 6
0 otherwise

6/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 Rb
(i) P(X < 2) = a
f (x)dx
R2 1
= −3
dx
6
1
= [x]2−3
6
1 5
= (2 + 3) =
6 6
(ii) P(|X | < 2) = P(−2 < X < 2)
R2 1
= −2 dx
6
1
= [x]2−2
6
4 2
= =
6 3
7/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 (iii)P(|X − 2| < 2) = P(−2 < x − 2 < 2)
P(−2 + 2 < x < 2 + 2)
P(0 < x < 4)
R3 1
= 0 dx
6
1 3
= [x]0
6
3 1
= =
6 2
(iv) P(X > k) = 1/3
R3
K
f (x)dx = 1/3
R3 1
dx = 1/3
K
6
1 3
[x] = 1/3
6 k 8/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 1
[3 − k] = 1/3
6
3−k =2
k =1

Problem: 2
Buses arrive at a specied stop at 15 minutes intervals
starting at 7 A.M., that is they arrive at 7, 7.15, 7.30,
7.45 and so on. If a passenger arrives at the stop at
a random time that is uniformly distributed between 7
and 7.30 A.M, nd the probability that he waits (i) less
than 5 minutes for a bus and (ii) at least 12 minutes
for a bus.
Solution:
Let X denote the time in minutes past 7 A.M, when the pas-
senger arrives at the stop.
Then X is uniformly distributed over(0, 30). 9/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 1
i.e., f (x) = , 0 < X < 30
30
(i) The passenger will have to wait less than 5 minutes if he
arrives at the stop between 7.10 and 7.15 or 7.25 and 7.30.
∴ Required probability = P(10 < X < 15) + P(25 < X < 30)
R 15 1 R 30 1
= 10 30 dx + 25 30 dx

1 1 30
= [x]15 10 + [x]
30 30 25
1 1 1
[5] + [5] =
30 30 3

10/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 (ii) The passenger will have to wait at least 12 minutes if he
arrives at the stop between 7.00 and 7.03 or 7.15 and 7.18.
∴ Required probability = P(0 < X < 3) + P(15 < X < 18)
R3 1 R 18 1
= 0 30 dx + 15 30 dx

1 1
= [x]30 + [x]18
30 30 15
1 1 1
= [3] + [3] =
30 30 5

HOME WORK
1. If X is uniformly distributed over (−α, α), α < 0, nd α
so that (i) P(X > 1) = 1/3, (ii) P(|X | < 1) = P(|X | > 1).

11/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 Gamma or Erlang Distribution
A continuous random variable x is said to have a Gamma dis-
tribution distribution if its probability density function is given
by

λk k−1 −λx
x e for x ≥ 0; k > 0



f (x) = Γ(k)

0

otherwise


Where Γ(k) is a Gamma function dened as

R∞
Γ(k) = 0 e −x x k−1

Derivation: To nd Moment Generating Function, Mean and

Variance of Gamma distribution
MX (t) = E {e tx }
R∞
= 0 e tx f (x)dx 12/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 R∞ λk k−1 −λx
= 0 e tx x e dx
Γ(k)
λk R ∞ −(λ−t)x k−1
= e x dx
Γ(k) 0
λk Γ(k) Γ(n) R ∞ −ax n−1
= (∵ = 0 e x dx )
Γ(k) (λ − t)k an
λk
=
(λ − t)k
λ
=[ ]k
(λ − t)
t
= [1 − ]−k
λ
d
Mean= µ1 = E (X ) = X̄ = [ [MX (t)]]t=0
0

dt
13/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 d t
=[ [1 − ]−k ]t=0
dt λ
t −1
= [−k[1 − ]−k−1 ( )]t=0
λ λ
k t −k−1
= [(1 − ) ]t=0 −→ (1)
λ λ
k
=
λ
Now
d2
= µ2 = E (X 2 ) = [ 2 [MX (t)]]t=0
0

dt
d k t −k−1
= [ [ [(1 − ) ]]t=0
dt λ λ
k t −1
= [(−k − 1)(1 − )−k−2 ( )]t=0
λ λ λ
k k +1
= [ ]
λ λ 14/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 k2 k
= 2
+ 2
λ λ
∴ Var (X ) = E (x 2 ) − [E (x)]2
k2 k k2
= 2 + 2 − ( 2)
λ λ λ
k
= 2
λ
For Gamma distribution

 λk k−1 −λx
x e for x ≥ 0; k > 0
f (x) = Γ(k)
0 otherwise


λ k k
MGF = MX (t) = [ ]k , Mean = , Variance = 2
(λ − t) λ λ
15/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 Problem: 1
In a certain city, the daily consumption of electric power
in millians of kilowatt hours can be treated as a random
variable having Erlang distribution with parameters λ =
1/2 and k =3. If the power plant of this city has a
daily capacity of 12 millians kilowatt-hours, what is the
probability that this power supply will be inadequate on
any given day.
Solution:
Let X be the the daily consumption of electric power in millians
of kilowatt hours.
Then the density function of X is given as
λk k−1 −λx
f (x) = x e for x ≥ 0; k > 0
Γ(k)
P (the power supply
R ∞is inadequate)
= P(X > 12) = 12 f (x)dx (∵ the daily capacity is only 12.)
16/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 R ∞ λk k−1 −λx
= 12 Γ(k) x e dx

R ∞ (1/2)3 2 −x/2
= 12 x e dx
Γ(3)
1 e −x/2 e −x/2 e −x/2 ∞
= [(x 2 )( ) − (2x)( ) + (2)( )]
8 × 2! −1/2 1/4 −1/8 12
1
= e −6 (288 + 96 + 16) = 25e −6 = 0.0625
16
Problem: 2
If a company employes n sales persons, its gross sales
in thousands of rupees may be regarded as a random
varaible having an Erlang
√ distribution with parameters
λ = 1/2 and k = 80 n. If the sales cost is Rs.8000
per sales person, how many sales persons should the
company employ to maximise the expected prot? 17/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 Solution:
Let X represents the gross sales (in Rupees) by n sales persons.
X follows√ the Erlang
√ distribution with√ parameters λ = 1/2 and
k = 80 n = 80 n × 8000 = 80000 n.
k √
∴ E (X ) = = 1, 60, 000 n
λ
If y denotes the total expected prot of the company, then
y =total expected sales − total sales cost
√
= 1, 60, 000 n − 8000n
dy 80000
= √ − 8000
dn n
√
= 0, when n = 10 or n = 100
d 2y 40000
2
= − 3/2 < 0, when n = 100.
dn n
∴, y is maximum when n = 100. That is the company should
employ 100 sales persons in order to maximise the total
18/ 40
expected prot. PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 Exponential Distribution
A continuous random variable x is said to follow an exponential
distribution with parameter λ > 0, if its probability density
function is given by

for x ≥ 0
 −λx
 λe
f (x) =
0 otherwise


Derivation: To nd Moment Generating Function, Mean and

Variance of Exponential distribution
MX (t) = E {e tx }
R∞
= 0 e tx f (x)dx
R∞
= 0 e tx λe −λx dx
R∞
= λ 0 e −(λ−t)x dx
19/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 e −(λ−t)x ∞
= λ[ ]
−(λ − t) 0
1
= λ( )
λ−t
λ
=
λ−t
d
Mean= µ1 = E (X ) = X̄ = [
0
[MX (t)]]t=0
dt
d
=[ [λ(λ − t)−1 ]]t=0
dt
= [λ(−1)(λ − t)−2 (−1)]]t=0
= [λ(λ − t)−2 ]t=0 −→ (1)
= λ(λ)−2
1
=
λ 20/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 Now
0 2 d2
µ2 = E (X ) = [ 2 [MX (t)]]t=0
dt
d
= [ [λ(λ − t)−2 ]]t=0
dt
= [λ(−2)(λ − t)−3 (−1)]t=0
= [2λ(λ − t)−3 ]t=0
2λ
=
λ3
2
= 2
λ
∴ Var (X ) = E (x 2 ) − [E (x)]2
2 1
= 2− 2
λ λ
1
= 2
λ 21/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 Memoryless Property of Exponential Distribution
Statement:
If X is exponentially distributed, then
P(X > s + t/X > s) = P(X > t) for any s, t > 0.
Proof
R∞
P(X > k) = k f (x)dx
R∞
= λ k e −λx dx
e −λx ∞
= λ[ ]
−λ k
∴ P(X > k) = e −λk −→ (1)
Now
P[(X > s + t) ∩ (X > s)]
P(X > s + t/X > s) =
P(X > s)
P(X > s + t)
=
P(X > s) 22/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 e −λ(s+t)
= by (1)
e −λs
e −λs e −λt
=
e −λs
= e −λt = P(X > t) by (1)
∴ P(X > s + t/X > s) = P(X > t).
Hence the proof.

Problem: 1
The time in hours is required to repair a machine is
exponentially distributed with parameter λ = 1/2. (i)
What is the probabilty that the required time exceeds
2 hours. (ii) What is the conditional probability that
a repaired time takes at least 10 hours. Given that its
duration exceeds 9 hours.
23/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 Solution:
Here λ = 1/2
Let X denote the time required to repair a machine.
W.K.T f (x) = λe −λx
(1) P (the repair time exceeds 2 hours)= P(X > 2)
R∞
= 2 f (x)dx
R∞
= 2 λe −λx dx
R∞
= 2 (1/2)e (−1/2)x dx
1 e (−1/2)x ∞
= [ ]
2 −1/2 2
= −(e −∞ − e −1 )
= e −1 = 0.3678
24/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 (ii) P(X > 10/X > 9) = P(X > 9 + 1/X > 9)

= P(X > 1)
= e −λt (By memoryless property)
= e (−1/2)1 = 0.6065

Problem: 2
The mileage which car owners get with a certain kind
of radial tire is a random variable having an exponential
distribution with mean 40000 km. Find the probabilty
that one of these tires will last (i) atleast 20000 km and
(ii) at most 30,000 km.
Solution:
Let X denote the mileage obtained with the cars.
X follows expnoential distribution.
25/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 Given mean = 40000
1
i.e; = 40000
λ
1
i.e; λ =
40000
W.K.T f (x) = λe −λx , x ≥ 0
−x
1
e 40000 dx
R∞
(i) P(X > 20000) = 20000
40000
1 e −x/40000 ∞
= [ ]
40000 −1/40000 20000
= e −0.5 = 0.6065
−x
R 30000 1
(ii) P(X ≤ 30000) = 0 e 40000 dx
40000
1 e −x/40000 30000
= [ ] = 1 − e −0.75 = 0.5270.
40000 −1/40000 0 26/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 Problem: 3
The life length of an electronic component follows an
exponential distribution. There are two processes by
which the component may be manufactured. The ex-
pected life length of component is 100 hrs, if process I
is used to manufacture, while it is 150 hours if process II
is used. The cost of manufacturing a single component
by process I is Rs.10, while it is Rs.20 for process II.
More over if the component lasts less than the guarn-
teed life of 200 hours , a loss of Rs 50 is to be borne
by the manufacturer, which process is advantageous to
the manufacturer?
Solution:
Let X denote the life length of an electronic component.
X follows exponential distribution.
W.K.T f (x) = λe −λx , x ≥ 0
27/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 If process I is used:
Given mean = 100
1
i.e; = 100
λ1
1
i.e; λ1 =
100
The life length of the component is 200 h. If before this
gaurantee time, the component is failed then Rs 50 will be
losed.
−x
R 200 1
P(X < 200) = 0 e 100 dx
100
1 e −x/100 200
= [ ]
100 −1/100 0
= 1 − e −2 = 0.8646
28/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 Similarly,
−x
R∞ 1
P(X ≥ 200) = 200 e 100 dx
100
1 e −x/100 ∞
= [ ] = e −2 = 0.1353.
100 −1/100 200
If process II is used:
Given mean = 150
1
i.e; = 150
λ2
1
i.e; λ2 =
150
Now,
−x
R 200 1
P(X < 200) = 0 e 150 dx
150
1 e −x/150 200
= [ ] = 1 − e −4/3 = 0.7364
150 −1/150 0 29/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 −x
R∞ 1
P(X ≥ 200) = 200 e 150 dx
150
1 e −x/150 ∞
= [ ] = e −4/3 = 0.2636.
150 −1/150 200
Let c1 and c2 be the costs per component corresponding to
the processes 1 and 2 respectively.

 10 for x ≥ 200


c1 =
60 for x < 200


 20 for x ≥ 200


c2 =
70 for x < 200


30/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 E (c1 ) = 10 × P(x ≥ 200) + 60 × P(x < 200)
= 10(.1353) + 60(.8646)
= 53.23
E (c2 ) = 20 × P(x ≥ 200) + 70 × P(x < 200)
= 10(.2636) + 60(.7364)
= 56.82
E (c1 ) < E (c2 )
Process I is advantageous to the manufacturer.

31/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 HOME WORK
1. The length shower in a tropical island in a rainy sea-
son has an exponential distribution with parameter 2,
time being measured in minutes. What is the probabil-
ity that it will last for at least one more minute.
2. The amount of time the watch will run without
having to be reset is a random variable having an ex-
ponential distribution with mean 120 days. Find the
probability that such a watch will (i) have to be set in
less than 24 days (ii) not have to be reset in atleast 180
days.

32/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 Weibull Distribution
A continuous random variable x has a Weibull distribution, if
its probability density function is given by

 αβx β−1 e −αx x ≥ 0;

β

for
f (x) =
0 elsewhere where α, β>0


NOTE: If β = 1, Weibull distribution is reduced to exponential

distribution.
Derivation: To nd Mean and Variance of Weibull distribution
R∞
E (x)= 0 xf (x)dx
R∞
= 0 xαβx β−1 e −αx −→ (1)
β

Put u = αx β
du = αβx β−1 dx 33/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 u
xβ = When x = 0, u = 0 and when x = ∞, u = ∞
α
1
u β
x =( )
α
1
R∞ u
(1) becomes E (x) = 0 ( ) β e −u du
α
1
1 R∞
= 1/β 0 u β e −u du
α
1
1 R ∞ −u β +1−1
= 1/β 0 e u du
α
1 1
= 1/β Γ(1 + )
α β
1 1
= 1/β ( )!.
α β
34/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 Now, R∞
E (x 2 ) = 0 x 2 f (x)dx
R∞
= 0 x 2 αβx β−1 e −αx dx −→ (1)
β

Put u = αx β
du = αβx β−1 dx
When x = 0, u = 0 and when x = ∞, u = ∞
1 2
u u u
x β = ⇒ x = ( ) β i.e; x 2 = ( ) β
α α α
2
R∞ u
(1) becomes E (x) = 0 ( ) β e −u du
α
2
1 R∞
= 2/β 0 u β e −u du
α
35/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 2
1 R∞ +1−1
= e −u u β du
α2/β 0
1 2
= 2/β Γ(1 + )
α β
1 2
= 2/β ( )!.
α β
∴ Var (X ) = E (x 2 ) − [E (x)]2
1 2 1 1
= 2/β Γ(1 + ) − [ 1/β Γ(1 + )]2
α β α β
1 2 1
Var(x) = 2/β [Γ(1 + ) − Γ2 (1 + )]
α β β

? Weibull distribution has no mgf.

36/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 Problem: 1
The life time of a component measured in hours is
Weibull distribution with parameters α = 0.2 and β =
0.5. Find (i) mean life time of the component (ii) the
probability that such a component will be more than
200 hours.
Solution LetX denote the life time of the component.
Given that α = 0.2 and β = 0.5
1 1
(i) Mean life time E (x) = 1/β Γ(1 + )
α β
1 1
= Γ(1 + )
(0.2)1/0.5 0.5
1
= Γ(3)
(0.2)2
2
= = 50
0.04 37/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 (ii) P {Life time greater than 200 hours}= P(X > 200)
R∞ 0.5
= 200 (0.2)(0.5)x −0.5 e −0.2x dx −→ (1)
Letu = 0.2x 0.5
du = (0.2)(0.5)x −0.5 dx
When x = 20, u = 2.8254 and When x = ∞, u = ∞
R∞
P(X > 200) = 2.8254 e −u du
= [−e −u ]∞
2.8254

= e −2.8254 = 0.059

Problem: 2
Find the probability of failure free performance of roller
bearings over a period of 104 hrs. If the life expectancy
of the bearings is depended by Weibull distribution with
parameters α = 10−7 and β = 1.5.
38/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 Solution
LetX denote the life time of the roller bearings in hours.
W.K.T f (x) = αβx β−1 e −αx for x ≥0
β

P (failure free)= P(X > 104 )

(i.e; if the life time is greater than 104 then failure will occur).
R∞
= 104 f (x)dx
R∞
= 104 αβx β−1 e −αx dx
β

−7 1.5
P(X > 104 ) = 104 10−7 (1.5)x 0.5 e −10 x dx
R∞
−→ (1)
Putu = −10−7 x 1.5
du = −10−7 (1.5)x 0.5 dx
1
When x = 104 =⇒ u = and when x = ∞, u = ∞
10
39/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G
 (1) becomes
R∞
P(X > 104 ) = 1/10 e
−u
du
= [e −u ]∞
1/10

= 0 + e −1/10 = 0.9

****** THANK YOU ******

40/ 40
PROBABILITY AND QUEUEING THEORY
By Dr.M.PERUMALSAMY ASSISTANT PROFESSORDEPARTMENT OF MATHEMATICS ALAGAPPA CHETTIAR G