You are on page 1of 40

Chapter 12: Queues and Process Simulation Modeling

2007 Pearson Education

Queueing Systems

Customer arrivals: people, machines, telephone calls, messages Servers: people, machines, airport runways, ATMs, computers Queue (waiting line): single, parallel, multiple with common line, series

2007 Pearson Education

Customer Characteristics

Arrival process: deterministic or probabilistic Poisson Process

customers arrive randomly and independently with mean rate l customers/time past does not influence the future probability of arrival does not vary over time (stationary)

Calling population: finite or infinite Behavior: reneging, balking, jockeying


2007 Pearson Education

Service Characteristics

Service process: deterministic or probabilistic Exponential services

mean service rate m customers/time (average service time is 1/m)

Number of servers: one or many

2007 Pearson Education

Queue Characteristics

Queue discipline: order in which customers are served FCFS LCFS Priority

2007 Pearson Education

System Configuration

One or more parallel servers fed by a single queue. Several parallel servers fed by their own queues. A combination of several queues in series.

2007 Pearson Education

Performance Measures

The quality of the service provided to the customer. Waiting time in the queue Time in the system (waiting time plus service time) Completion by a deadline The efficiency of the service operation and the cost of providing the service. Average queue length Average number of customers in the system (queue plus in service) Throughput -- the rate at which customers are served Server utilization -- percentage of time servers are busy Percentage of customers who balk or renege
2007 Pearson Education

Operating Characteristics

Lq = average number in the queue L = average number in the system Wq = average waiting time in the queue W = average time in the system P0 = probability that the system is
empty

2007 Pearson Education

Analytical Models Single Server Model

Assumptions Single server Poisson arrivals, mean rate = l Exponential services, mean rate = m FCFS queue discipline Other models Arbitrary service times Multiple servers Finite calling populations
2007 Pearson Education

Single Server Model Operating Characteristics


Average number in queue = Lq = l2/[mml
Average number in system = L = lml Average waiting time in queue = Wq = lmml Average time in system = W = 1/(ml Probability system is empty = 1 - lm
2007 Pearson Education

Example

Customers arrive at an airline ticket counter at a rate of l = 2 customers/minute and can be served at a rate of m = 3 customers per minute.

2007 Pearson Education

Calculations

Lq = L=

22 3(2)

=1.33 customers

2 33

= 2.00 customers

Wq =
W=

2 3(3 2)

= 0.67 minutes

1 3 2

= 1.00 minutes

P0 = 1 2/3 = 0.33
2007 Pearson Education

Analytical Models vs. Simulation

Analytical models provide only longterm steady-state results Simulation results show short-rerm transient behavior

2007 Pearson Education

Littles Law

For any steady-state queuing system, L = lW Other relationships Lq = lWq L = Lq + lm W = Wq + 1/m

2007 Pearson Education

Process Simulation Concepts


Customer arrives

Customer waits for service if server is busy


Customer receives service
next!

Customer leaves
2007 Pearson Education

Observations

If a customer arrives and the server is idle, then service can begin immediately upon arrival. If the server is busy when a customer arrives, then the customer cannot begin service until the previous customer has completed service. The time that a customer completes service equals the time service begins plus the actual service time.
2007 Pearson Education

Manual Process Simulation

2007 Pearson Education

Process Simulation with SimQuick

SimQuick Elements Entrances where objects enter a process. Buffers places where objects can be stored

(inventory storage, queues of people or parts, and so on). Work Stations places where work is performed on objects (machines, service personnel, and so on). Decision Points where an object goes in one of two or more directions (outcomes of processing activities, routings for further processing, and so on). Exits places where objects leave a process according to a specified schedule.
2007 Pearson Education

Process Simulation with SimQuick

Statistical Distributions Normal: Nor(mean, standard

deviation)

Exponential: Exp(mean) Uniform: Uni(lower, upper) Constant Discrete: Dis(i), where i is the reference to table i of the worksheet

2007 Pearson Education

SimQuick Control Panel

2007 Pearson Education

SimQuick Queuing Simulation Model

Customers at a car wash arrive randomly at an average of 15 cars per hour (or one car every 4 minutes). A car takes an average of 3 minutes to wash (or 20 cars per hour) Process flow map:

2007 Pearson Education

Entrances Worksheet

2007 Pearson Education

Buffers Worksheet

2007 Pearson Education

Work Stations Worksheet

2007 Pearson Education

Simulation Results

2007 Pearson Education

Work Station Statistics


Final status: status of the work station when the simulation ends Final inventory (int. buff.), Mean inventory (int. buff.), and Mean cycle time (int. buff.): Work cycles started: the number of times the work station has started processing Fraction time working: utilization of the work station Fraction time blocked: fraction of time that the work station was waiting to pass on an object to the next element.
2007 Pearson Education

Buffer Statistics

Objects leaving: number of objects that left the buffer Final inventory: Inventory refers to the number of objects in the buffer. Final inventory is the number remaining at the end of the simulation Minimum inventory; Maximum inventory; Mean inventory: statistics on the number of objects during the simulation Mean cycle time: mean time that an object spends in the buffer
2007 Pearson Education

Comparison to Analytical Results

2007 Pearson Education

Queues in Series with Blocking

2007 Pearson Education

Buffers Worksheet with Queue Capacities

2007 Pearson Education

Grocery Store Model with Resources

2007 Pearson Education

Resources Worksheets

2007 Pearson Education

Inspection Model with Decision Points

2007 Pearson Education

Decision Point Table

2007 Pearson Education

Pull System Supply Chain With Exit Schedules

2007 Pearson Education

Other SimQuick Features


Discrete distributions Custom schedules

2007 Pearson Education

Continuous Simulation Modeling

A continuous simulation model defines equations for relationships among state variables so that the dynamic behavior of the system over time can be studied.

2007 Pearson Education

Example: Cost of Medical Services

2007 Pearson Education

Modeling Equations

POPLVL(t) = POPLVL(t - 1) + GROWTH(t) DEMAND(t) = POPLVL(t) - [MEDRATE(t - 1) - MEDRATE(t - 2)] MEDRATE(t) = MEDRATE(t = 1) + POPLVL(t) - POPLVL(t - 1)

+ .8*[INSRATE(t - 1) - INSRATE(t - 2)]


INSRATE(t) = INSRATE(t = 1) + .10*MEDSUIT(t - 1) - [RISK(t - 1) RISK(t - 2)] MEDSUIT(t) = MEDSUIT(t - 1) + [MEDRATE(t - 1) - 1]/RISK(t - 1) RISK(t) = RISK(t - 1) + .10*[MEDSUIT(t - 1) - 1]

2007 Pearson Education

Simulation Results

2007 Pearson Education

You might also like