Fish Swarming Behaviour

This is part of the TU Delft Research Project. It serves as a way to recreate the results described in the paper Protection algorithms using fault resilient fish swarming behaviour.

Libraries

pygame : pip install pygame

colorsys : pip install colorsys

---

Prey Predator

To run this instance:

• With custom weights:

python preyPredator/preyPredator.py -b {FAULTY_BOIDS} -p {FAULTY_PREDATORS}

• With the default weights:

python preyPredator/preyPredator.py

To use the genetic algorithm for weight training:

python script/training_script.py

---

Missile Target

• With custom weights:

python preyPredator/missileTarget.py -b {FLEE_WEIGHT} -p {PROTECT_WEIGHT}

• With the default weights:

python preyPredator/missileTarget.py

---

Results

To generate the graphs for each instance:

• Run the runner script for that instance:

python scrips/preyPredator_runner.py

python scrips/rocketTarget_runner.py

• Run the plotting script for that instance:

python plotting/plotPreyPredator.py

python plotting/plotRocketTarget.py

• The results are firstly generated in the results folder and then plotted

---

Other remarks

• All weights and constants can be modified in the constants file of each instance
• Training instances can also be run individually using:

python preyPredator/geneticsTraining.py