A Q-learning based arithmetic optimization algorithm for a multi-warehouse joint replenishment and delivery problem

The joint replenishment and delivery (JRD) strategy is critical for increasing operational management efficiency and reducing costs. This study introduces a new multi-warehouse JRD model for heterogeneous products. The main goal of the JRD strategy is to minimize costs by determining the optimal replenishment intervals, delivery frequency, and basic replenishment cycle time. To address the difficulties of the JRD optimization issue, we propose the Q-learning-based arithmetic optimization algorithm (QAOA). In the QAOA framework, Q-learning is the guiding principle, making decisions based on current conditions and constantly refining search strategy via feedback mechanisms. Furthermore, an escape mechanism has been implemented to reduce the possibility of algorithmic stagnation in local optima. Experiments show that QAOA exceeds eight popular benchmark algorithms. By employing the QAOA technique, the practical JRD model has been effectively handled, resulting in considerable cost reductions in supply chain management.