Robust optimization of high-speed railway train plan based on multiple demand scenarios

A train plan usually needs to be implemented on multiple operation days, while there are certain differences in the total amount and space-time distribution of the demand of each operation day. To enable the same train plan implemented on multiple operation days to effectively adapt different demands of multiple operation days, robust optimization of high-speed railway train plan based on multiple demand scenarios was studied gradually. Firstly, the demand distribution on each operation day was described as a demand scenario. A candidate-train-set was generated for adapting the multiple demand scenarios. Then analyzed train operation costs and passenger travel costs, the regret value of total costs of the two was used to measure the deviation between the object value generated by the robust scheme and the optimal object value under each demand scenario, and a robust optimization model of high-speed railway train plan was constructed, which aimed at minimizing the maximum regret value. Based on neighborhood solution search strategies for multiple demand scenarios, a simulated annealing algorithm for solving the model was designed. Finally, take Ning-Hang high-speed railway line as an example for analysis, which showed that the robust scheme was only 0.56% away from the target value of implementing satisfactory scheme respectively on each operation day. And if randomly select one of the satisfactory schemes of all demand scenarios to be implemented on all operation days, their overall costs are larger than the robust scheme, so it can effectively adapt time-related demand of multiple operation days. It is proved that the proposed robust optimization method for train plan has certain validity and feasibility. © 2022, Central South University Press. All rights reserved.