Treffer: A Column Generation-Based Optimization Approach for the Train Loading Planning Problem with Simulation-Based Evaluation of Rail Forwarding at the Port of Valencia.

As ports evolve to meet sustainability targets, seamless coordination between road and rail operations becomes fundamental to success. This study addresses the Train Loading Planning Problem (TLPP) which focuses on assigning outbound containers to train wagons under slot, weight, and pattern constraints aiming to examine its broader systemic implications. A compact mixed-integer programming formulation is developed and enhanced through a column-generation approach that efficiently prices feasible wagon plans. The optimization module is embedded within a discrete-event simulation of terminal processes including yard handling, gate operations, and train timetables. The study tests a TLPP-based rail planning algorithm within a DES of terminal and hinterland operations to quantify the impact under realistic variability. Using operational data from the Port of Valencia, realistic planning scenarios are evaluated across varying demand mixes and train frequencies. Results indicate that integrating rail capacity with optimized wagon loading reduces set-up time by 20%, delivery lead time by 54%, container dwell time by 80%, and greenhouse gas emissions by 54% compared with a trucking forwarding baseline, while maintaining throughput and alleviating congestion at terminal gates and yards. From a computational perspective, the column-generation approach achieves improved runtimes to the compact MIP and scales linearly to the number of variables. The proposed framework delivers ready to use load plans and practical insights for the deployment of additional rail capacity, supporting sustainable logistics in port environments. [ABSTRACT FROM AUTHOR]

Copyright of Future Transportation is the property of MDPI and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)