A Decomposition-Based Evolutionary Algorithm with Adaptive Weight Vectors for Multi- and Many-objective Optimization

The multi-objective evolutionary algorithms based on decomposition (MOEA/D) have achieved great success in the area of evolutionary multi-objective optimization. Numerous MOEA/D variants are focused on solving the normalized multi- and many-objective problems without paying attention to problems having objectives with different scales. For this purpose, this paper proposes a decomposition-based evolutionary algorithm with adaptive weight vectors (DBEA-AWV) for both the normalized and scaled multi- and many-objective problems. In the light of this direction, we compare existing popular decomposition approaches and choose the best suitable one incorporated into DBEA-AWV. Moreover, one novel replacement strategy is adopted to attain the balance between convergence and diversity for multi- and many-objective optimization problems. Our experimental results demonstrate that the proposed algorithm is efficient and reliable for dealing with different normalized and scaled problems, outperforming several other state-of-the-art multi- and many-objective evolutionary algorithms.