Stratified opposition-based initialization for variable-length chromosome shortest path problem evolutionary algorithms

摘要: Abstract Initialization is the first and a major step in implementation of evolutionary algorithms (EAs). Although there are many common general methods to initialize EAs such as pseudo-random number generator (PRNG), no single method that can fit every problem. This study provides new, flexible, diversity-aware, easy-to-implement initialization for genetic algorithm shortest path The proposed algorithm, called stratified opposition-based sampling (SOBS), considers phenotype genotype diversity while striving achieve best fitness population. SOBS does not depend on specific type sampling, because main goal stratify space. aims at an initial population with higher genotype. To investigate performance SOBS, four network models were used simulate real-world networks. Compared most frequently method, is, PRNG, more accurate solutions, better running time less memory usage, fitness. Statistical analysis showed yields solutions accuracy 68–100% time. this was focused it be applied other population-based solve problem use same direct representation particle swarm optimization (PSO).