Evolutionary Adjustment of Probabilistic Models based on Cellular Automata applied to Fire Propagation

Forest fires have increased as a result of climate change, causing severe environmental damage. Cellular Automata (CA) models can simulate fire spread, aiding decision-making, but their parameter tuning is complex. This research develops a Genetic Algorithm (GA) to automate CA parameter adjustment. The study evaluates different fitness functions and GA’s ability to reproduce the fire dynamics present in datasets produced under varying sampling intervals. The experimental results show that GA effectively tunes the parameters, achieving accurate fire propagation across models.