Modification of Genetic Algorithm Based on Extinction Events and Migration

This article presents a genetic algorithm modification inspired by events related to great extinctions. The main objective of the was minimize number function solutions until minimum for established. It assumed that, within each step, population should be smaller than that recommended in applicable literature, iterations limited, solution area variable, and extinction event take place following several iterations. Calculations were performed 10 individuals population, iterations, two generations each, with happening once every three developed presented, capable indicating Eggholder Rastrigin functions, higher probability master (default “ga” MATLAB) at same solutions. An proposed focusing on minimizing randomization function, which may an alternative surrogate model. Typically, emphasis is achieving as much accuracy possible. method obtaining highest possible accuracy. A presented minimizes disadvantages largest computation time need generate many samples typical algorithms (GAs). Optimization results classic GA, GEGA, WOA, SMA, SSA functions compared. made obtain global extreme satisfactory sufficiently high probability, while calculated basis function. methodology used fulfill target turbine disc.