Analysis of inter and intra-front operations in multi-modal multi-objective optimization problems

Abstract Many real-world multi-objective optimization problems inherently have multiple multi-modal solutions and it is in fact very important to capture as many of these possible. Several crowding distance methods been developed the past few years approximate optimal solution search space. In this paper, we discuss some shortcomings distance-based such inaccurate estimates density neighboring We propose a new classification for selection operations Pareto-based algorithms. This based on utilizing nearby from other fronts calculate values. Moreover, address drawbacks existing methods, two algorithms whose mechanisms are each introduced types operations. These called NxEMMO ES-EMMO. Our proposed evaluated 14 test various complexity levels. According our results, most cases, algorithm with mechanism produces more diverse space comparison competitive