An Enhanced Artificial Bee Colony Optimization Algorithm

摘要

Artificial Bee Colony (ABC) algorithm has depicted more robust convergence than various bio-inspired optimization algorithms on numerous benchmark functions. Moreover, ABC algorithm has capability to replace the exhausted food-sources with newly picked food-sources. Nonetheless, ABC has shown to suffer from slow convergence and poor exploitation. Additionally, the new inducted food-sources as a replacement of an exhausted food-source generally-have poor quality and hence, the algorithm converges slowly. Various solutions have been proposed to overcome the demerits of ABC, but so far the variants are either computationally intensive or could not avert the flaws. This research work proposes two different modifications for the performance enhancement of ABC algorithm. The first modification enhances neighborhood searching capability and the second modification improves new-food-source inducting capability of ABC algorithm. The proposed algorithm has been compared with various existing variants of ABC algorithm on a few benchmark functions. The significance of the proposed algorithm has been verified using a statistical test. The result-analysis reveals that the proposed algorithm has resulted in the best convergence among all compared ABC variants.