Non-deterministic events and processes

Discrete Event Simulation in Python

Diogo Costa (PhD, MSc)

Adjunct Professor, University of Saskatchewan, Canada & CEO of ImpactBLUE-Scientific

Event non-determinism

NOTE: Terms "event" and "process" used interchangeably.

What is event (or process) non-determinism?

Inability of process to be predicted accurately
Process is referred to as "Non-Deterministic"

Why are deterministic processes important?

Natural and human-driven activities often involve non-deterministic processes
Every time they occur, they repeat in different way
Their duration cannot be determined accurately
May have strong impact on the system output

Examples of non-deterministic processes or events

Natural world

Moment a volcano will erupt
Exact moment it will rain
Time and location lighting will strike
Time and location a tornado will form

Human-driven or human-initiated processes

Moment machine will break
Moment one feels the need to go to bathroom
Moment baby starts crying
If commercial flight will be delayed

Representing non-deterministic events in discrete-event models

Non-deterministic processes

Produce non-repeatable impact on system
Unknown exact duration

Statistics

Represent variability in duration of processes

The random package is helpful for this
Example: random.randint(start, end)

import random

# Non-deterministic processes
process_1 = [5, 15]

while time < simulation_time:

   # Update simulation time: process_1
   time += random.randint(
               process_1[0], 
               process_1[1])

Representing non-deterministic events using SimPy

No specific methods to account for non-deterministic processes
Event duration variability calculated as for non-SimPy-based models
Pass right statistics to .timeout()

import random

# Non-deterministic processes
process_1 = [5, 15]

while True:

  # Update simulation time: process_1
  env.timeout(random.randint(
         process_1[0], 
         process_1[1]))

Let's practice!

Discrete Event Simulation in Python