Meta-Reinforcement Learning via Evolution for Multi-Objective Combinatorial Supply Chain Optimisation
For researchers and practitioners in supply chain optimization and multi-objective RL, this method improves Pareto front diversity and adaptation, though it is an incremental extension of existing meta-RL and evolutionary techniques.
This paper tackles multi-objective combinatorial supply chain optimization using meta-reinforcement learning. The proposed population-based framework with evolutionary search achieves up to 32% higher hypervolume and the lowest average Hausdorff distance compared to existing methods.
Meta-reinforcement learning is a promising approach to multi-objective optimisation because it enables rapid policy adaptation across changing environments and preference settings. However, conventional few-shot methods usually fine-tune from a single shared meta-policy, which can reduce solution diversity and limit exploration of the Pareto front, especially in high-dimensional combinatorial problems such as supply chain optimisation. We propose a population-based Meta-reinforcement learning framework that combines decomposition with evolutionary search in scalarisation weight space. The framework maintains a population of weight vectors, each associated with a distinct meta-policy trained through gradient-based meta-learning, and iteratively refines this population through elitist selection, crossover, and mutation guided by hypervolume and entropy contributions. We evaluate the method in a multi-objective supply chain setting with conflicting economic, environmental, and social goals, and further test its generality on standard reinforcement learning problems. The results show that the proposed approach yields more diverse, better distributed Pareto front approximations, improves cross-task adaptation, increases hypervolume by up to 32\% over Meta-multi-objective reinforcement learning in the complex case, and attains the lowest average Hausdorff distance among all compared methods.