Optimization model and method for existing railway alignment reconstruction

With the substantial growth of railway traffic demand, to enhance the transport capacity of the railway network, in addition to new railways, numerous low-grade existing railways need to be reconstructed. At present, the reconstruction of existing railways largely depends on manual design. Thus, to improve the quality and efficiency of reconstruction design, it is urgent to study the intelligent reconstruction of existing railway alignments. However, most of the existing researches on alignment optimization focus on new railway alignments, without considering the complex coupling relations between existing and reconstructed alignments. Thus, these methods cannot be directly applied to the optimization design of existing railway alignment reconstruction. For the optimization model and method for existing railway alignment reconstruction, an optimization model was developed by defining the geometric parameters of the automatic segmentation results as design variables. The reconstruction cost was formulated as an objective function, and three categories of constraints (geometric, structural and location constraints) were considered. To solve this model, a two-stage method was developed. Firstly, an automatic segmentation mechanism for sequential screening was proposed for segmenting existing alignments into reused and reconstructed sections. Then, a multi-directional scanning strategy was devised and incorporated into distance transform to search for redesigned alignments. By being applied to a real-world case, the optimized solution can save 7.2% of the reconstruction cost compared with the best manual solution. The effectiveness of the proposed method is verified through data analysis. © 2023, Central South University Press. All rights reserved.