Side-Blotched Lizard Algorithm: A polymorphic population approach

Abstract

In metaheuristic algorithms, finding the optimal balance between exploration and exploitation is a key research topic that remains open. In the nature, a reptile called Side-Blotched Lizard has achieved an interesting dynamic balance over its population. Such lizards evolved with three morphs associated with distinctive mating strategies. The synergy between the morphs generates a polymorphic population, able to self-balance the subpopulations of each morph without depleting the weakest morph. This equilibrium is achieved as the most common morph becomes the weakest, and the smaller subpopulations increase their chances of mating. In this paper, the Side-Blotched Lizard Algorithm (SBLA) is proposed to emulate the polymorphic population of the lizard. For this purpose, three operators are used to guarantee a dynamic over the population that allows the coexistence of multiple morphs. From the computational point, SBLA uses a subpopulation managing strategy which emulates the sinusoidal distribution of the lizard population over time. Even more, the mating behavior of each morph is modeled with three concepts, defensive, expansive, and sneaky. The performance of SBLA is tested on a set of five unimodal, eighteen multimodal, four composite benchmark functions, and engineering problems like; the welded beam, FM synthesizer, and rolling element bearing. To validate the results, we compared them to ten well-established algorithms and using the Wilcoxon test and the Bonferroni correction. The examination of the experimental results exhibits the accuracy, robustness and unique problem-solving method of the proposed algorithm.