A Systematic Study on Recent Evolutionary Algorithms for Multi-Modal Multi-Objective Optimization

Abstract

The real-world optimization problems are inherently multi-modal and multi-objective (MMMO), such as bio-medical imaging, automotive engine design, plant identification in control systems, inference engine design, etc. This is mainly because of the acute diversity and diffusion of solutions in Pareto space and the multi-modality of the solution set. Therefore, finding the optimal solution for MMMO is a pressing need in the literature. It is evident from the literature that evolutionary Algorithms (EA) are the best candidates for solving MO problems. However, due to massive variants of single-objective EAs (SOEA) and Multi-objective EA (MOEA), the advocacy of the best EA for the MMMO problem needs to be improved in the literature. This work has presented a comprehensive study on the recent variants of EA to solve single-objective and multi-objective optimization problems. In addition, the Uni-modal and multi-modal problems are also considered with the respective EA.