Anais do Simpósio Brasileiro de Pesquisa Operacional
Anais do Simpósio Brasileiro de Pesquisa Operacional

This paper studies a novel Multi-Objective Evolutionary Algorithm~(MOEA), the Non-dominated Sorting Biased Random-Key Genetic Algorithm~(NS-BRKGA), designed for Multi-objective Optimization Problems~(MOPs).
The NS-BRKGA integrates elements from both the Multi-Parent Biased Random-Key Genetic Algorithm with Implicit Path-Relinking~(BRKGA-MP-IPR) and the Elitist Non-dominated Sorting Genetic Algorithm~(NSGA-II).
By managing the population diversity within the variable space, it enhances solution quality and diverges from traditional approaches.
The paper details a comparative study conducted using the ZDT test suite for continuous bi-objective problems and includes several other algorithms, such as NSPSO, MOEA/D-DE, MHACO, and IHS.
Despite the NSGA-II emerging as the most effective algorithm, the NS-BRKGA demonstrated competitive performance, underlining its potential applicability to continuous MOPs even though it was primarily designed for combinatorial problems.
Further research could focus on better understanding and refining the NS-BRKGA's performance for continuous problems.