In multiobjective evolutionary algorithms based on decomposition (MOEA/D), how to generate offspring significantly influences the convergence speed and the population diversity. In this paper, we propose an offspring generation strategy for MOEA/D. In this strategy, we firstly generate multiple temporary solutions for each subproblem in MOEA/D, secondly employ the whole parent population to construct surrogate model, thirdly, employ surrogate model to sort and identify the optimal temporary solution, and finally, generate offspring by employing differential evolution (DE) operator with polynomial mutation on the optimal temporary solution. We conduct extensive experiments and compare our method with some state-of-the-art multiobjective evolutionary algorithms. The experimental results show that our method obtains the best performance.
An Offspring Generation Strategy for MOEA/D
01.10.2019
114779 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
MOEA/D based UAV swarm deployment for wireless coverage
IEEE | 2021
|Optimization Design of Distributed Propeller Position Based on MOEA/D Algorithm
Springer Verlag | 2024
|STUDIES ON THE APPLICATION OF MOEA/D TO AERODYNAMIC OPTIMIZATION DESIGN
British Library Conference Proceedings | 2018
|Audi A6, offspring of the new strategy
Online Contents | 2011
|