- I have used an algorithm close to GA with turbines as the individuals. The dna is a dictionary which contains the coordinates.
- If speed loss is less than average loss, that turbine is killed and a new turbine is mutated by crossover.The crossover is a random weighted sum of the two parents and i know, its illogical to spawn based on coordinates ; lets try :P
- I have used a primitive genetic algorithm with turbines as the individuals. The dna shape is lifespan * [x, y]
- At every timestep of life, speed loss is calculated first and fitness is calculated based on speed loss
- After a generation ends, selection is done based on a threshold on fitness.
- Less-fit turbines are mutated by a single-point crossover of two randomly selected turbines
#install dependencies
pip3 install -r requirements.txt
#run algorithms : <algorithm> = v1, v2
python3 <algorithm>/game.py