Adaptive Multi-Strategy Artificial Bee Colony Algorithm Based on Dynamic Dimension Disturbance

In order to enhance the search performance of artificial bee colony algorithm in dealing with complex optimization problems, this paper proposes an adaptive multi-strategy artificial bee colony algorithm (LMDABC) based on dynamic dimension perturbation. In LMDABC, Latin hypercube initialization method is introduced to replace the random method in basic ABC in the initialization stage to achieve the goal of uniform distribution of initial population in space. At the same time, a variety of search strategies are used and an adaptive mechanism is introduced to meet the needs of different iterations. In addition, the dynamic dimension perturbation method is used to generate more potential optimal solutions to improve the convergence performance of the algorithm. The performance of LMDABC is verified by relevant experiments with CEC2014 test set. From the results, LMDABC performs well on most functions.