Multi-strategy boosted mutative whale-inspired optimization approaches

This paper presents an improved Whale Optimization Algorithm (WOA) for global optimization. WOA is a recently introduced meta-heuristic algorithm mimicking the hunting behavior of humpback whales. Owing to its simplicity in exploratory and exploitative operators and the satisfactory efficacy, this algorithm has found its place among the well-established population-based approach utilized in many engineering and science areas. However, this method is easy to fall into local optimum when dealing with some optimization cases. In order to further enhance its exploratory and exploitative performance, three strategies are incorporated into the original method to keep a better balance between exploitation and exploration tendencies. First, the chaotic initialization phase is introduced into the optimizer to initiate the swarm of chaos-triggered whales. Then, Gaussian mutation is employed to intensify the diversity level of the evolving population. At last, a chaotic local search with a 'shrinking' strategy is used to enhance the exploitative leanings of the basic optimizer. In order to verify the effectiveness of the improved WOA, it is compared to four meta-heuristic and state-of-the-art evolutionary algorithms on representative benchmark functions. Trial results and simulations reveal that not only the proposed improved WOA is significantly better than those basic algorithms including original WOA but also it is superior to compared state-of-the-art approaches. Moreover, the proposed algorithm is successfully applied to realize three constrained engineering test cases, which the results suggest that the improved WOA can effectively deal with the constrained functions as well. (C) 2019 Elsevier Inc. All rights reserved.