Improving Artificial Bee Colony Algorithm with Historical Archive.

In this study, an artificial bee colony with historical archive (HAABC) is proposed to help ABC escape from stagnation situation. The proposed framework keeps track of the search history and stores excellent successful solutions into an archive. Once stagnation is detected in scout bees phase of ABC, a new individual is generated by utilizing the historical archive. Experimental results on 28 benchmark functions show that the proposed framework significantly improves the performance of basic ABC and five state-of-the-art ABC algorithms.