Integrated optimization of transfer station selection and train timetables for road-rail intermodal transport network

In this study, we investigate the integrated optimization of container transfer station selection and train timetables for a road-rail intermodal transport network based on the transport organization mode of improved passenger-like container trains. The aim is to achieve a prompt response from train timetables for freight transportation needs to improve the quality of the timetable. The problem is formulated as a mixed-integer nonlinear programming problem, and a multiobjective bi-level optimization model is proposed with the objectives of increasing the average utilization ratio of the rail capacity and decreasing the number of road-rail transfer stations, transport costs, and transport delay penalty costs. To improve the quality of the Pareto solutions, a grid-based adaptive artificial bee colony (GBA-ABC) algorithm is designed to solve the model of a road-rail network with multicommodity flows. As a case study, the model is applied to the high-speed railway network in China and the China-Europe transcontinental railway network. The results show that the integrated optimization approach yields train timetables that can simultaneously reduce the transportation costs and the transportation delay penalty costs, with a rail capacity utilization rate of above 79.57%. Concurrently, it is beneficial for solving robust solutions under different case scales, demonstrating the effectiveness of the model and the algorithm. Finally, the proposed algorithm is experimentally compared with other multiobjective metaheuristics algorithms and is found to exhibit superior performance.

Automated summary

This study investigates the integrated optimization of container transfer station selection and train timetables in a road-rail intermodal transport network. The aim is to improve rail capacity utilization and reduce transportation costs. Using a grid-based adaptive artificial bee colony algorithm, effective solutions are achieved for different case scales.