Research on Optimization of Station Shunting Operation Scheduling Based on Resource Availability

This study focuses on the optimization of station shunting operation scheduling under resource availability constraints and proposes a multi-agent collaborative optimization method based on the Deep Q-Network (DQN). The study first analyzes resource constraints (shunting locomotives, tracks, personnel) and operational scenarios, constructs a resource state model, then designs a multi-dimensional state space and dynamic action space, integrates the DQN algorithm to achieve optimal decision-making, and avoids route conflicts through a collaboration mechanism. Case verification shows that this method can output optimized marshalling and pickup/delivery operation plans, improve shunting efficiency and resource utilization, and provide new ideas for intelligent railway operation scheduling.