Enhancing artificial bee colony algorithm with generalised opposition-based learning

As a new global optimisation technique, artificial bee colony (ABC) algorithm becomes popular in recent years for its simplicity and effectiveness. In the basic ABC, however, the solution search equation updates only one dimension to produce a new candidate solution, which may result in that the offspring becomes similar to its parent and cause insufficient search. To overcome this drawback, we proposes an enhanced ABC (EABC) variant by utilising the generalised opposition-based learning (GOBL) strategy. With the help of GOBL, much more promising search regions can be explored, so the probability of converging to the global optimum is highly increased. Experiments are conducted on 13 well-known benchmark functions to verify the proposed approach, and the results show that EABC is very promising in terms of solution accuracy and convergence speed.