A Mating Selection Based on Modified Strengthened Dominance Relation for NSGA-III

In multi/many-objective evolutionary algorithms (MOEAs), to alleviate the degraded convergence pressure of Pareto dominance with increase in number objectives, numerous modified relationships were proposed. Recently, strengthened relation (SDR) has been proposed, where area a solution is determined by degree and niche size (?¯). Later, controlled SDR (CSDR), ?¯ an additional parameter (k) associated are dynamically adjusted depending on iteration count. Depending problem characteristics distribution current population, different situations require values k, rendering linear reduction k based generation count ineffective. This because particular value expected bias relationship towards region front (PF). addition, due same reason, using or CSDR environmental selection cannot preserve diversity solutions required cover entire PF. Therefore, we propose MOEA, referred as NSGA-III*, (1) (MSDR)-based mating adaptive ensemble would prioritize parents from specific sections PF (2) traditional weight vector non-dominated sorting-based NSGA-III protect corresponding The performance NSGA-III* favourably compared state-of-the-art MOEAs DTLZ WFG test suites up 10 objectives.