Stochastic Scheduling of High-Speed Railway Meal Service for On-Time Delivery

This study focuses on the scheduling of high-speed railway (HSR) meal service for on-time delivery. Different from ordinary takeout services, meal service of HSR has a tight requirement on delivery time, and is expected to arrive exactly at its due date (i.e., the departure time of a train). Such difference causes a lot of challenges in providing meal  service for HSR. One challenge is the simultaneous consideration of two conflict objectives that have very different structures, earliness cost and tardiness cost. Another challenge is the randomness in the system, which significantly increases the computational burden of objective evaluation. To tackle these challenges, the mathematical program of the problem is formulated to minimize the expected earliness and tardy job costs first. Theoretical analysis is further conducted to reveal the effects of a few key factors on the optimal policy and the costs. The considered problem can be transformed into a two-stage stochastic mixed-binary program, and a surrogate algorithm is proposed to solve the problem. Extensive numerical experiments on a real-world case study are conducted to demonstrate the effectiveness of the model and the proposed algorithm. Experiment results show that the proposed algorithm possesses superior efficiency and acceptable accuracy compared with the commercial solver. Moreover, the necessity of considering processing time uncertainties is also manifested via a series of comparison experiments. These results suggest that both the considered model and the proposed solution approach are helpful for practitioners to make high-quality HSR scheduling decisions in reality.

Language

  • English

Media Info

  • Media Type: Digital/other
  • Features: Figures; Maps; References; Tables;
  • Pagination: 22p

Subject/Index Terms

Filing Info

  • Accession Number: 01908115
  • Record Type: Publication
  • Report/Paper Numbers: TRBAM-24-00020
  • Files: TRIS, TRB
  • Created Date: Feb 14 2024 2:33PM