Description
PART II
Write an object-oriented program to simulate the operation of an emergency department at a
small, rural hospital.
The emergency department has four areas to handle patients. All patients initially visit the
Triage area so that their cases can be prioritized based on clinical need. Some patients may be
discharged from the Triage area (and the emergency department) if they can be treated very
quickly. Patients who cannot be treated quickly will move from the Triage area to one of the
other areas in the emergency department before being discharged. The Trauma area handles
the most seriously ill or injured patients. The Acute Care area handles patients who are
seriously ill but not in immediate danger. Patients with conditions that are not life-threatening
are seen in the Prompt Care area.
The program shall implement a discrete-event simulation with a simulation clock using the
next-event time advance method. Each service area shall have a FIFO queue with one or more
servers (medical specialists). Configuration data for the simulation stop time and each service
area shall be read from a text file. The path to the input file shall be a command line parameter.
Input file format:
[simulation stop time]
[number of servers] [triage inter-arrival mean] [triage service mean] [triage discharge probability]
[number of servers] [trauma care service mean] [trauma care probability]
[number of servers] [acute care service mean] [acute care probability]
[number of servers] [prompt care service mean] [prompt care probability]
Example input file data:
10080
1 6.0 7.0 0.10
1 180.0 0.10
1 85.0 0.30
3 15.0 0.50
An explanation of the example input file data follows.
The simulation shall end after 10,080 minutes (7 days).
Triage area
1 triage specialist
Inter-arrival time from exponential distribution with mean of 6 minutes.
Service time from exponential distribution with mean of 7 minutes.
Patients will be discharged from the emergency department with probability of 0.10.
Trauma Care area
1 trauma care specialist
Service time: mean of 180 minutes
Patients will move from Triage to Trauma Care with probability of 0.10.
Acute Care area
1 acute care specialist
Service time: mean of 85 minutes
Patients will move from Triage to Acute Care with probability of 0.30.
Prompt Care area
3 prompt care specialists
Service time: mean of 15 minutes
Patients will move from Triage to Prompt Care with probability of 0.50.
Output
The output printed to the console shall include:
• The total number of patients discharged from the Triage area who do not require additional care.
• The total number of patients discharged from Trauma Care, Acute Care, and Prompt Care areas.
• Report the average delay in queue(s) for patients.
• The average number of patients in queue over time for each queue
• Server utilization over time for each server in each care area.
You can implement your assignment using Java, C++, or Python. The program shall create at
least two object types: a server object type and a patient object type. These object types shall
be defined in separate files.
Submit your source code files and any required build script using the directory structure below.
hw2/
[build script]
[input file: ed.txt]
src/
[source files]
Upload your submission to Blackboard as a zipped file: lastname-firstname.hw2.zip.
INFORMATION FROM IN CLASS DISCUSSION
Info on exponential distribution can be found on Page 28 in textbook.
Most of the logic needed can be found in Example 1.4 in textbook.
Also, check the flowcharts in the textbook for arrival, departure,
and other events for proper modeling.
T = -B*ln(u(0,1)) where -B is -lambda
Max Queue size should be 100 patients. If we get beyond that,
terminate program and print why. Perhaps make Queue size a
command line parameter.
Mr. Apostal’s Class structure for PART I:
Customer/Patient
ArrivalTime
DelayInQueue
DepartureTime ?
Server
BusyStatus
QueueOfPatients
SysClock
SimTime
NextArrivalTime
NextDepartTime
Driver
(Drives the program)
For Part II:
Look for the average number of patients in all queues over time.
All patients must be processed through the Triage queue.
“Server Utilization” is the % of time each server was busy during the simulation.
INFORMATION FROM BLACKBOARD ANNOUNCEMENTS
Program #1 Note:
The generated random numbers for inter-arrival and service times should
be floats or doubles (32- or 64-bit–this may depend on the language you use).
Report metrics at the end of the program should be rounded to 3 decimal positions.
Program #1, Additional Note:
If the number of patients in the triage area (queue) is greater than or equal to 100,
your simulation should stop. Your program should then print a message stating why the
simulation was stopped (ex “number of patients in queue exceeded pre-defined threshold”).
In this case the program should not print any report metrics. Also, the threshold value
should be a command-line argument.
Program #1, Generating Exponential Random Numbers:
To generate random variates (random variables) from an exponential distribution,
see the 5th edition of the text, section 1.4.3, bottom half of p. 28. You can also
read in the Wikipedia article, Exponential distribution, the section on Generating
Exponential Variates. The symbol beta (β) is sometimes used to refer to the mean
time of an exponential distribution. The symbol lambda (λ) = 1 / β
