| Abstrakt: |
In the field of multi-objective optimization algorithms (MOAs), a primary challenge is identifying an optimal solution that effectively converges to the true Pareto Front while maintaining a high level of diversity, especially in the initial phases of evolution. To tackle this challenge, this study introduces a novel algorithm named Many-Objective Dragonfly Algorithm (MaODA). This method incorporates reference point, niche preserve and information feedback mechanism for maintaining better balance between convergence and diversity. Through this study, the MaODA effectiveness is established by using the WFG1-WFG9 test suite for 4, 6 and 8 objectives and five real-world (RWMaOP1- RWMaOP5) design problems. The findings demonstrate that MaODA outperforms its peer many-objective algorithms' like MaOTLBO, MaOJAYA, MaOPSO and NSGA-III in most test scenarios in terms of GD, IGD, SP, SD, HV and RT metrices. It is shown that MaODA achieves better convergence, as well as enhanced diversity and spread, compared to peers. [ABSTRACT FROM AUTHOR] |
Full text from SpringerLink