Dynamic migration in multiple ant colonies

Multi-population-based bio-inspired computation may use migration among groups to increase the search diversity. Through good solutions exchanged among sub-populations, better solutions may be found with a high probability. In this paper, we propose two algorithms to dynamically adjust the two primary parameters, migration interval and migration rate, for flexibly reflect solution situation for effective migration. The first algorithm only dynamically changes the migration interval, and the second considers both interval and rate. We will examine how the dynamic migration strategies affect the quality of solutions in the experiments.