Modelling and optimization of constrained alignments for existing railway reconstruction
Hao Pu, Huidan Fu, Paul Schonfeld, Taoran Song, Lu Zhao, Xianbao Peng
Abstract
Presently, numerous low-grade existing railways need to be reconstructed to improve their capacity. The core of the supporting analysis process redesigns railway alignments based on existing ones. However, most alignment optimization studies focus on new alignment designs but none have considered the complicated constraints between existing and redesigned alignments. This paper focuses on optimizing constrained alignments for existing railway reconstruction. An optimization model is developed by defining existing alignments into reused and reconstructed sections. The reconstruction cost is formulated as an objective function and three categories of constraints are considered. To solve this model, a two-stage method is developed. Firstly, an automatic segmenting method is proposed for separating existing alignments into the aforementioned two kinds of sections. Then, a multi-directional distance transform is devised to search for redesigned alignments. By being applied to a real-world case, the effectiveness of the proposed method is verified through data and sensitivity analyses.