Dados

=Tas de llegada del cleinte (1/ = Tiempo promedio de

arribo)
=Tasa de sevicio (1/ = Tiempo promedio de servicio)
s =Numero de servidores
Calculate
L
q
=Numero promedio de clientes en la cola
L =Numero promedio de clientes en el sistema
W
q
=Tiempo promedio de espera en la cola
W =Tiempo promedio en el sistema (incluyendo el servicio)
P
n
=Probabilidad de tener n clientes en el sistema
=Utilizacin del sistema
BasicRelationships
Las sisguientes relaciones son para todos los modelos.
La utilizacin del sistema es
El numero promedio de clientes en el sistema es
El tiempo promedio de espera en la cola es
El tiempo promedio de espera en el sistema es(incluyendo el servicio) is

s
L L
q
+

W
q

L
q

W W
q
+
1

Model1(M/M/1)
Formulas
Probabilidad de que el sistema este vacio:
Probabilidad de n clientes en el sistema:
Numero promedio en la cola:
Ejemplo
Elreadeatencinausuariosdelabbliotecarecibesolicitudesde
atencinenunpromediode10porhora(DistribucinPoisson).
Existeslounbibliotecarioenestarea,ypuedeatenderalos
cleintesenunpromediode5minutos(DistribucinExponential).
Calessonlasmedidasdefuncionamientodelsistema?
P
0
1

P
n
P
0

_
,

n
L
q

2
( )
Template for M/M/s Queueing Model
Data Results
10 (mean arrival rate) L = 5
12 (mean service rate)
L
q
=
4.166666667
s = 1 (#servers)
W = 0.5
Pr(w>t) =
0.1353353
W
q
=
0.416666667
when t = 1
r = 0.833333333
Prob( w
q
>t) =
0.1127794
when t = 1
P
0
=
0.166666667
P
1
=
0.138888889
P
2
=
0.115740741
P
3
=
0.096450617
P
4
=
0.080375514
P
5
=
0.066979595

Model 2 (M/G/1)
Formulas
Numero promedio en lnea:
Probabilidad que el sistema este vacio:
(Caso especial: M/D/1)
Ejemplo
EllavadodeautosABCesunlavadoautomtico.Cadaclientemanejasuauto
alreadelavado,lodejayesperamientraselcarroesautomaticamente
lavado.Losautoslleganaleatoriamenteaunpromediode20autosporhora.
Eltiempodeservicioesexactamentede2minutos.Cualessonlasmedidasde
funcionamiento?
L
q

2
+
2
2 1 ( )
P
0
1

L
q

2
2 ( )
Template for M/D/1 Queueing Model
Data Results
20 (mean arrival rate) L = 1.33333333
30 (mean service rate)
L
q
=
0.66666667
s = 1 (# servers)
W = 0.06666667
W
q
=
0.03333333
0.66666667
P
0
=
0.33333333
Model 3 (M/M/s)
Formulas
Probabilidad de que el sistema este vacio:
Probabilidad de n clientes en el sistema:
Probabilidad que un nuevo arribo tenga que esperar:
Numero promedio en la cola:
P
0

1
/ ( )
n
n!
+
/ ( )
s
s!(1 )
n0
s1

P
n

( / )
n
n!
P
0
for n 1K , s
/ ( )
n
s!s
ns
P
0
for n s

'





P
w

_
,

s
P
0
s!(1 )
L
q
P
0
( / )
s+1
(s 1)!(s / )
2



1
]
1
Ejemplo
El Banatec de Tepic tiene tres cajeros en su sucursal Tecnolgico. Los
clientes llegan aleatoriamente a la sucursal a una tasa de 1 por minuto. La
tasa de servicio es en promedio de 2 minutos, y sigue una distribucin
exponencial?
Template for M/M/s Queueing Model
Data Results
60 (mean arrival rate) L = 2.888888889
30 (mean service rate)
L
q
=
0.888888889
s = 3 (# servers)
W = 0.048148148
Pr(>t) =
1.341E-12
W
q
=
0.014814815
when t = 1
0.666666667
Prob(
q
>t) =
4.159E-14
when t = 1
P
0
=
0.111111111
P
1
=
0.222222222
P
2
=
0.222222222
P
3
=
0.148148148
P
4
=
0.098765432
P
5
=
0.065843621
Aplicaciones de Teora de Colas
Podemos usar los resultados de la
teora de colas para realizar los
siguientes tipos de decisiones:

Cuantos servidores
emplear

Usar un servidor rpiso o

varios servidores lentos

Usar servidores de uso

general o especfico
Meta:
Minimizar el costo total = costo de los servidores+ costo de espera


Cost of
Service Capacity
Cost of customers
waiting
Total Cost
Optimum
Service Capacity
Cost
Ejemplo #1: Cuntos servidores?
En el departamento de servicio de un taller de reparacin de autombiles, los
mecnicos que requieren partes para reparaciones presentan sus formas de requisicin
al almacen. El dependiente entrega el pedido mientras espera el mecnico. Los
mecnicos llegan a una tasa promedio de 40 por hora (poisson). El dependiente entrega
el pedido en 3 minutos (exponential). Si al dependiente se le paga a $6 por hora y
almecnico $18 por hora, cual es el numero ptimo de dependientes en el almacen.
Template for Economic Analysis of M/M/s Queueing Model
Data Results
40 (mean arrival rate) L = 2.88889
20 (mean service rate)
L
q
=
0.88889
s = 3 (# servers)
W = 0.07222
Pr(>t) =
0.0012726
W
q
=
0.02222
when t = 1
0.66666667
Prob(
q
>t) =
0.0008484
when t = 1
P
0
=
0.11111
P
1
=
0.14815
C
s
=
$6 (cost/server/unit time)
P
2
=
0.09877
C
w
=
$18 (waiting cost/unit time)
P
3
=
0.06584
Cost of Service = $18.00
P
4
=
0.04390
Cost of Waiting = $52.00
P
5
=
0.02926
Total Cost = $70.00
P
6
=
0.01951
Template for Economic Analysis of M/M/s Queueing Model
Data Results
40 (mean arrival rate) L = 2.17391
20 (mean service rate)
L
q
=
0.17391
s = 4 (# servers)
W = 0.05435
Pr(>t) =
4.54E-05
W
q
=
0.00435
when t = 1
0.5
Prob(
q
>t) =
2.27E-05
when t = 1
P
0
=
0.13043
P
1
=
0.17391
C
s
=
$6 (cost/server/unit time)
P
2
=
0.08696
C
w
=
$18 (waiting cost/unit time)
P
3
=
0.04348
Cost of Service = $24.00
P
4
=
0.02174
Cost of Waiting = $39.13
P
5
=
0.01087
Total Cost = $63.13
P
6
=
0.00543
Template for Economic Analysis of M/M/s Queueing Model
Data Results
40 (mean arrival rate) L = 2.03980
20 (mean service rate)
L
q
=
0.03980
s = 5 (# servers)
W = 0.05100
Pr(>t) =
6.144E-06
W
q
=
0.00100
when t = 1
0.4
Prob(
q
>t) =
2.458E-06
when t = 1
P
0
=
0.13433
P
1
=
0.17910
C
s
=
$6 (cost/server/unit time)
P
2
=
0.08955
C
w
=
$18 (waiting cost/unit time)
P
3
=
0.03582
Cost of Service = $30.00
P
4
=
0.01433
Cost of Waiting = $36.72
P
5
=
0.00573
Total Cost = $66.72
P
6
=
0.00229
Por lo que s = 4 tiene el menor costo total.
Example #2: How Many Servers?
Beefy Burgers is trying to decide how many registers to have open
during their busiest time, the lunch hour. Customers arrive during the
lunch hour at a rate of 98 customers per hour (Poisson distribution).
Each service takes an average of 3 minutes (exponential distribution).
Management would not like the average customer to wait longer than
five minutes in line. How many registers should they open?
Template for M/M/s Queueing Model
Data Results
98 (mean arrival rate) L = 51.4655
20 (mean service rate)
L
q
=
46.5655
s = 5 (# servers)
W = 0.5252
Pr(>t) =
0.1429016
W
q
=
0.4752
when t = 1
0.9800
Prob(
q
>t) =
0.1286114
when t = 1
P
0
=
0.0008
Template for M/M/s Queueing Model
Data Results
98 (mean arrival rate) L = 7.3593
20 (mean service rate)
L
q
=
2.4593
s = 6 (# servers)
W = 0.0751
Pr(>t) =
1.19E-08
W
q
=
0.0251
when t = 1
0.8167
Prob(
q
>t) =
1.54E-10
when t = 1
P
0
=
0.0053
Choose s = 6 since W = 0.0751 hour is less than 5 minutes.
Example #3: One Fast Server or Many Slow Servers?
Beefy Burgers is considering changing the way that they serve
customers. For most of the day (all but their lunch hour), they have
three registers open. Customers arrive at an average rate of 50 per
hour. Each cashier takes the customers order, collects the money, and
then gets the burgers and pours the drinks. This takes an average of 3
minutes per customer (exponential distribution). They are considering
having just one cash register. While one person takes the order and
collects the money, another will pour the drinks and another will get
the burgers. The three together think they can serve a customer in an
average of 1 minute. Should they switch to one register?
3 Slow Servers
Template for M/M/s Queueing Model
Data Results
50 (mean arrival rate) L = 6.0112
20 (mean service rate)
L
q
=
3.5112
s = 3 (# servers)
W = 0.1202
Pr(>t) =
6.376E-05
W
q
=
0.0702
when t = 1
0.8333
Prob(
q
>t) =
3.188E-05
when t = 1
P
0
=
0.0449
1 Fast Server
Template for M/M/s Queueing Model
Data Results
50 (mean arrival rate) L = 5.0000
60 (mean service rate)
L
q
=
4.1667
s = 1 (# servers)
W = 0.1000
Pr(>t) =
4.54E-05
W
q
=
0.0833
when t = 1
0.8333
Prob(
q
>t) =
3.783E-05
when t = 1
P
0
=
0.1667
W is less for one fast server, so choose this option.
Example #4: General or Specific Servers?
A small bank in a mall has two tellers. One handles only merchant
customers and one handles regular customers. Merchant customers
and regular customers each arrive at an average rate of 20 per hour
(for a total arrival rate of 40 customers per hour). The service time
for both tellers averages 2 minutes (exponential). The bank manager
is considering changing the setup to allow each teller to handle both
merchant customers and regular customers. Since the tellers would
have to handle both types of jobs, their efficiency would decrease to
a mean service time of 2.2 minutes. Should they switch to the new
setup?
Specific Tellers (1 for Merchants, 1 for Regular)
Template for M/M/s Queueing Model
Data Results
20 (mean arrival rate) L = 2.0000
30 (mean service rate)
L
q
=
1.3333
s = 1 (# servers)
W = 0.1000
Pr(>t) =
4.54E-05
W
q
=
0.0667
when t = 1
0.6667
Prob(
q
>t) =
3.027E-05
when t = 1
P
0
=
0.3333
General Tellers (2 Tellers Handle Both Jobs)
Template for M/M/s Queueing Model
Data Results
40 (mean arrival rate) L = 3.1731
27.2727 (mean service rate)
L
q
=
1.7064
s = 2 (# servers)
W = 0.0793
Pr(>t) =
6.408E-07
W
q
=
0.0427
when t = 1
0.7333
Prob(
q
>t) =
2.991E-07
when t = 1
P
0
=
0.1538
Waiting times are less when both tellers handle both jobs, so choose this option.