Biogeography-based optimization (BBO) is a novel evolutionary algorithm, which is proposed with inspiration from the science of biogeography, to solve global optimization problems. To overcome the inability of BBO to make a good balance between its exploration and exploitation abilities, this paper introduces a new variant of BBO, which is called NBBO. The framework of NBBO considers two or more sub-iterations in an iteration of the algorithm to perform the evolution process. In each sub-iteration, a sample (sub-population) is selected from the input population of the iteration, based on a triangular probability distribution, to choose emigrating habitats (solutions) from. On the other hand, a novel two-phase migration operator is used in the framework of NBBO to make the algorithm effectively explore a search space. By making a good balance between its exploration and exploitation abilities, which is conducted by its new framework, NBBO can escape from local optima. Quantitative evaluations, based on extensive experiments on a set of 23 benchmark functions with diverse complexities, reveal that NBBO achieves favorable results which are quite superior to the results of other relevant state-of-the-art swarm intelligence-based and evolutionary algorithms.
