Demand clustering-based design of circular train services for freight railway transportation

The circular train circulates fixed railcars to serve designated freight stations in a seamless loop. This type of arrangement not only elevates logistics service standards, but also enhances resource utilization. This paper introduces a demand-driven framework tailored for designing circular train services, leveraging comprehensive loading data from the China Railway (CR) network. This framework initiates a clustering analysis to identify potential freight demands for circular train services within the planning horizon. Subsequently, we transform the origin and destination (OD) pairs of those identified demands into a network of interconnected nodes. An optimization model rooted in the Multiple Traveling Salesman Problem (MTSP) is presented for optimizing circular train services. A case study and comparative experiments based on historical data from CR demonstrates the working process of the proposed method and verify its performance. The results show that the organization of circular train services can achieves a 6% reduction in empty transportation compared to the overall CR freight network.