Reinforcement Learning-Based Hybrid Multi-Objective Optimization Algorithm Design

The multi-objective optimization (MOO) of complex systems remains a challenging task in engineering domains. methodological approach applying MOO algorithms to simulation-enabled models has established itself as standard. Despite increasing computational power, the effectiveness and efficiency such algorithms, i.e., their ability identify many Pareto-optimal solutions possible with few simulation samples possible, plays decisive role. However, question which class is most effective or efficient respect problems not yet been resolved. To tackle this performance problem, hybrid that combine multiple elementary search strategies have proposed. potential, no systematic for selecting combining Pareto suggested. In paper, we propose an designing uses reinforcement learning (RL) techniques train intelligent agent dynamically strategies. We present both fundamental RL-Based Hybrid (RLhybMOO) methodology exemplary implementation applied mathematical test functions. results indicate significant gain agents over static strategies, highlighting effectively efficiently select algorithms.