Genetic Algorithm with Immigration Like Strategies of Diversification

This paper presents an improvement in the process of diversification of populations in the genetic algorithms using immigration and an appropriate selection of operators. The proposal significantly enhance the quality of solutions obtained, because it prevents premature convergence like result from the loss of diversity in the genetic material of individuals by inbreeding between them, the inbreeding can be avoided with the migration operator. The developed algorithm was run with the instances kroA100, and gil269 of the benchmarks available in the TSPLIB and some results are shows in this work.