Efficient and merged biogeography-based optimization algorithm for global optimization problems

In order to improve the optimization efficiency of the biogeography-based optimization (BBO) algorithm, this study proposes a novel BBO algorithm, namely an efficient and merged biogeography-based optimization (EMBBO) algorithm. Firstly, BBO’s mutation operator is got rid of. Then, a differential mutation operator and a sharing operator are merged into BBO’s migration operator to obtain an improved migration operator. In the improved migration operator, the emigration habitats are selected by a new example learning approach. The above improvements can enhance the optimization performance and reduce the computation complexity. Thirdly, a new single-dimensional and all-dimensional alternating strategy is combined with the improved migration operator to balance exploration and exploitation and reduce more computation complexity. Fourthly, the opposition-based learning approach is merged to prevent the algorithm from falling into the local optima. Finally, the greedy selection method is used instead of the elitist strategy to avoid setting the elitist parameter and to get rid of one sorting step. We make a large number of experiments on a set of classic benchmark functions and CEC2017 test set and apply EMBBO to clustering optimization. Experiment results verify that EMBBO can obtain the highest optimization efficiency compared with quite a few state-of-the-art algorithms.