Basic Models
AnyLogic
Author: Alejandro RodrIguez Grau
Tutor: Prof. Dr. Angel A. Juan
Full Course Index
• Introduction to AnyLogic
• Basic models
• Heterogeneous model
• Advanced model
Basic Models
• Exercise 1
Access Control Point
Overview of the problem
Create the Model
Run and Results
• Exercise 2
Check-in Point
Overview of the problem
Create the Model
Run and Results
1. A simple access control point
Process overview: Passengers arrive to a passport
control point. They wait in line and show their passports
to any of two agents. Then, passengers move to the
boarding card control. After validating their cards, they
can proceed to the boarding door.
Assumptions: (a) Passengers interarrival times follow
an Exponential(0.5) minutes; (b) the time an agent
takes to interact with the passenger follows a
Triangular(0.4, 0.9, 1.5) minutes; (c) the time the airline
staff member takes to interact with the passenger
follows a Triangular (0.3, 0.4, 0.6) minutes; (d) the travel
time between each process is 15 seconds; (e)
passengers will wait in a single waiting line on a FIFO
basis; (f) there is no limit to the length of the waiting
lines.
2. Create a new Model, Add Agent Type
This model will work best if each
step represents 1 minute inside
the simulation.
Drag & Drop an Agent Type from
the Agent Pallete.
Add Parameters and
Variables on new
Agent Type:
Passenger
[Link] & Variables for optional use
We could define a parameter to fix the
value of the time in a path.
But in our case, we will use connectors into Our Variables will be used, as in the first
a Delay block, where Entities will wait those model, to obtain and print results. So for
15 virtual seconds, before proceeding to now, lets go back to the Main tab, and
the next step. create the proposed model.
4. Implement your Model (1/2)
Drag & Drop the Source, Services (2) and Sink, change
their names accordingly and set up the defined parameters
on the first slide of this exercise
4. Implement your Model (2/2)
5. Add Delay (1/2)
Now we already have a simulation to run, entities
have an arrival rate, processing time on the server
and a sink to close the cycle.
Still, entities are instantly going from the “out”
nodes to the “in” nodes. We must add a Path or a
Delay, where we can specify the time that it takes
for a passenger to go from one place to another (in
this case, a fixed time).
5. Add Delay (2/2)
Once defined the waiting time, we must set a high
enough capacity, because if we leave only 1 (by
default), no other entity will be able to do that path
until the one in it reaches the exit.
Do this for every Delay block
6. Use Parameters in Main (1/2)
Now that we know how to input data, lets learn
how to substitute a number that appears several
times and can by compressed into a Parameter
inside the “Main” Agent screen.
6. Use Parameters in Main (2/2)
Use that parameter instead of the numbers in
each Delay block.
Remember to
change the
Agent Type
on the Delay
blocks so the
program can
find its way to
the
parameter!
7. Add Results
Time to revisit the
Reorder parameters and variables according to
Passenger Agent.
what is needed, some can go in Main, others
must go in the Agent in question.
Notice that it’s no longer as seen in slide number
5, this is not unusual, as when you program, you
sometimes find better ways or improvements for
the system performance along the way.
8. Show Results on console (1/2)
8. Show Results on console (2/2)
9. Add other simulations
Add an Experiment to
your model workspace.
Adding a new experiment is basically adding a new
mode of simulation for the program.
We can change things as randomness, repetition,
speed, or others present in the properties when
selecting the new simulation on Projects.
10. A simple check in point
Process overview: People arrive to an airport and go
through the check in process to get their tickets. After
check-in the passengers proceed to the security check
point.
Assumptions: (a) Passengers interarrival times follow
an Exp(1) min.; (b) passengers walk rate follows a
Uniform(2, 4) km/h; (c) passengers travel 50 m from the
terminal entrance to the check in station; (d) following
the check in process they must walk 65 m to the
security check point; (e) the check-in station has 4
people assigned to process customers, who wait in a
FIFO single line; (f) the check-in process follows a
Uniform(2, 5) to completion; (g) the simulation model
needs to be run for 24h.
11. Create the model
Follow the same steps as before, Drag &
Drop objects into the model, connect them
via a connector.
We will use Delay blocks to account for the
time the passenger stays on the path.
12. Agent Passenger
We create our Agent
called Passenger, and Don’t forget the metrics you are using! (This
we add variables and example shows the number for a simulation
parameters that we need which each step equals 1 second)
to show results.
13. First simulation
We can see how the simulation behaves, and we
must check that it is not doing anything funny, like
agents instantly skipping a block, travelling to fast,
or not generating enough entities at the source.
With the simulation running we can set new
parameters/variables to get data from the
simulation on console.
14. Adding results
We must search for the Agent Actions on each block and
use this spaces to order the program to produce results, it
will be very helpful if you look first the example model, so
you can see how this coding part looks like.