15.1 Some random selection problems (Page 6/6)

Applied probability Page 6 / 6

Remark . The counting process is a Poisson process in the sense that $N_{t} \sim$ Poisson $(λ t)$ for all $t > 0$ . More advanced treatments show that the process has independent, stationary increments. That is

$N (t + h) - N (t) = N (h)$ for all $t, h > 0$ , and
For $t_{1} < t_{2} \leq t_{3} < t_{4} \leq \dots \leq t_{m - 1} < t_{m}$ , the class ${N (t_{2}) - N (N_{1}), N (t_{4}) - N (t_{3}), \dots, N (t_{m}) - N (t_{m - 1})}$ is independent.

In words, the number of arrivals in any time interval depends upon the length of the interval and not its location in time, and the numbers of arrivals in nonoverlappingtime intervals are independent.

Emergency calls

Emergency calls arrive at a police switchboard with interarrival times (in hours) exponential (15). Thus, the average interarrival time is 1/15 hour (four minutes).What is the probability the number of calls in an eight hour shift is no more than 100, 120, 140?

p = 1 - cpoisson(8*15,[101 121 141])p = 0.0347 0.5243 0.9669 Got questions? Get instant answers now!

We develop next a simple computational result for arrival processes for which $S_{n} \sim$ gamma $(n, λ)$ .

Gamma arrival times

Suppose the arrival times $S_{n} \sim$ gamma $(n, λ)$ and g is such that

\int_{0}^{\infty} | g | < \infty and E [\sum_{n = 1}^{\infty}, | g (S_{n}) |] < \infty

Then

E [\sum_{n = 1}^{\infty}, g, (S_{n})] = λ \int_{0}^{\infty} g

VERIFICATION

We use the countable sums property (E8b) for expectation and the corresponding property for integrals to assert

E [\sum_{n = 1}^{\infty}, g, (S_{n})] = \sum_{n = 1}^{\infty} E [g (S_{n})] = \sum_{n = 1}^{\infty} \int_{0}^{\infty} g (t) f_{n} (t) d t where f_{n} (t) = \frac{λ e^{- λ t} {(λ t)}^{n - 1}}{(n - 1)!}

We may apply (E8b) to assert

\sum_{n = 1}^{\infty} \int_{0}^{\infty} g f_{n} = \int_{0}^{\infty} g \sum_{n = 1}^{\infty} f_{n}

Since

\sum_{n = 1}^{\infty} f_{n} (t) = λ e^{- λ t} \sum_{n = 1}^{\infty} \frac{{(λ t)}^{n - 1}}{(n - 1)!} = λ e^{- λ t} e^{λ t} = λ

the proposition is established.

Got questions? Get instant answers now!

Discounted replacement costs

A critical unit in a production system has life duration exponential $(λ)$ . Upon failure the unit is replaced immediately by a similar unit. Units fail independently. Cost of replacement of a unit is c dollars. If money is discounted at a rate α , then a dollar spent t units of time in the future has a current value $e^{- α t}$ . If S _n is the time of replacement of the n th unit, then $S_{n} \sim$ gamma $(n, λ)$ and the present value of all future replacements is

C = \sum_{n = 1}^{\infty} c e^{- α S_{n}}

The expected replacement cost is

E [C] = E [\sum_{n = 1}^{\infty}, g, (S_{n})] where g (t) = c e^{- α t}

Hence

E [C] = λ \int_{0}^{\infty} c e^{- α t} d t = \frac{λ c}{α}

Suppose unit replacement cost $c = 1200$ , average time (in years) to failure $1 / λ = 1 / 4$ , and the discount rate per year $α = 0.08$ (eight percent). Then

E [C] = \frac{1200 \cdot 4}{0.08} = 60, 000

Got questions? Get instant answers now!

Random costs

Suppose the cost of the n th replacement in [link] is a random quantity C _n , with ${C_{n}, S_{n}}$ independent and $E [C_{n}] = c$ , invariant with n . Then

E [C] = E [\sum_{n = 1}^{\infty}, C_{n}, e^{- α S_{n}}] = \sum_{n = 1}^{\infty} E [C_{n}] E [e^{- α S_{n}}] = \sum_{n = 1}^{\infty} c E [e^{- α S_{n}}] = \frac{λ c}{α}

Got questions? Get instant answers now!

The analysis to this point assumes the process will continue endlessly into the future. Often, it is desirable to plan for a specific, finite period. The result of [link] may be modified easily to account for a finite period, often referred to as a finite horizon .

Finite horizon

Under the conditions assumed in [link] , above, let N _t be the counting random variable for arrivals in the interval $(0, t]$ .

If Z_{t} = \sum_{n = 1}^{N_{t}} g (S_{n}), then E [Z_{t}] = λ \int_{0}^{t} g (u) d u

VERIFICATION

Since $N_{t} \geq n$ iff $S_{n} \leq t$ , $\sum_{n = 1}^{N_{t}} g (S_{n}) = \sum_{n = 0}^{\infty} I_{(0, t]} (S_{n}) g (S_{n})$ . In the result of [link] , replace g by $I_{(0, t]} g$ and note that

\int_{0}^{\infty} I_{(0, t]} (u) g (u) d u = \int_{0}^{t} g (u) d u

Got questions? Get instant answers now!

Replacement costs, finite horizon

Under the conditions of [link] , consider the replacement costs over a two-year period.

SOLUTION

E [C] = λ c \int_{0}^{t} e^{- α u} d u = \frac{λ c}{α} (1 - e^{- α t})

Thus, the expected cost for the infinite horizon $λ c / α$ is reduced by the factor $1 - e^{- α t}$ . For $t = 2$ and the numbers in [link] , the reduction factor is $1 - e^{- 0.16} = 0.1479$ to give $E [C] = 60000 \cdot 0.1479 = 8, 871.37$ .

Got questions? Get instant answers now!

In the important special case that $g (u) = c e^{- α u}$ , the expression for $E [\sum_{n = 1}^{\infty}, g, (S_{n})]$ may be put into a form which does not require the interarrival times to be exponential.

General interarrival, exponential g

Suppose $S_{0} = 0$ and $S_{n} = \sum_{i = 1}^{n} W_{i}$ , where ${W_{i} : 1 \leq i}$ is iid. Let ${V_{n} : 1 \leq n}$ be a class such that each $E [V_{n}] = c$ and each pair ${V_{n}, S_{n}}$ is independent. Then for $α > 0$

E [C] = E [\sum_{n = 1}^{\infty}, V_{n}, e^{- α S_{n}}] = c \cdot \frac{M_{W} (- α)}{1 - M_{W} (- α)}

where M _W is the moment generating function for W .

DERIVATION

First we note that

E [V_{n} e^{- α S_{n}}] = c M_{S_{n}} (- α) = c M_{W}^{n} (- α)

Hence, by properties of expectation and the geometric series

E [C] = c \sum_{n = 1}^{\infty} M_{W}^{n} (- α) = \frac{M_{W} (- α)}{1 - M_{W} (- α)}, provided | M_{W} (- α) | < 1

Since $α > 0$ and $W > 0$ , we have $0 < e^{- α W} < 1$ , so that $M_{W} (- α) = E [e^{- α W}] < 1$ .

Got questions? Get instant answers now!

Uniformly distributed interarrival times

Suppose each $W_{i} \sim$ uniform $(a, b)$ . Then (see Appendix C ),

M_{W} (- α) = \frac{e^{- a α} - e^{- b α}}{α (b - a)} so that E [C] = c \cdot \frac{e^{- a α} - e^{- b α}}{α (b - a) - [e^{- a α} - e^{- b α}]}

Let $a = 1$ , $b = 5$ , $c = 100$ and $α = 0.08$ . Then,

a = 1;
b = 5;c = 100;
A = 0.08;MW = (exp(-a*A) - exp(-b*A))/(A*(b - a))
MW =    0.7900EC = c*MW/(1 - MW)
EC =  376.1643

Got questions? Get instant answers now!

<< Chapter < Page Page > Chapter >>

Read also:

Get Jobilize Job Search Mobile App in your pocket Now!

100% Free Mobile Applications
Receive real-time job alerts and never miss the right job again

Source: OpenStax, Applied probability. OpenStax CNX. Aug 31, 2009 Download for free at http://cnx.org/content/col10708/1.6

Google Play and the Google Play logo are trademarks of Google Inc.

Notification Switch

Would you like to follow the 'Applied probability' conversation and receive update notifications?

Ask

	1 BOD - DERMATOLOGY - By Brooke Delaney Start Exam
	Back Safety Quiz - "Back in Action" By David Martin Start Quiz
	16 AP 16 Neurological Essay Exam By OpenStax Start Flashcards
	4 Biology 04 Cell Structure MCQ By OpenStax Start Quiz
	2 Timeshare 2 By Jams Kalo Start Quiz
©flickr: Gage	How well do you know Ross Lynch? By Brianna Beck Start Quiz
	Cultural Anthropology Assignment 2 By Richley Crapo Start Assignment
	1 AP 01 Human Body Anatomy Physiology MCQ By OpenStax Start Quiz
©flickr: Luis	Atoms By Carly Allen Start Quiz
	Professional Writing MCQ By Mary Cohen Start Quiz