Variable Neighborhood Search for Multistage Train Classification at Strategic Planning Level. (30th August 2017)

• Record Type: Journal Article
• Authors: Belošević, Ivan; Ivić, Miloš
• Is Part Of: Computer-aided civil and infrastructure engineering. Volume 33:Number 3(2018:Mar.)
• Page Start: 220
• Page End: 242
• Publication Date: 2017-08-30
• Subjects: Civil engineering -- Data processing -- Periodicals; Computer-aided engineering -- Periodicals; 624.0285
• DOI: 10.1111/mice.12304
• Languages: English
• ISSNs: 1093-9687

Abstract

Abstract: Considerations on influences and interrelations between structural design and functional planning are imperative for raising the cost and operational effectiveness of marshaling yards in the context of managing wagonload services. This article focuses on multistage train classification, which rearranges a set of wagons from inbound trains to create a set of multigroup outbound trains in a proper form. Previous papers on the subject of multistage train classification were limited to considerations on the optimal classification schedule addressing the number of sorting steps as a primary objective. In contrast, this article proposes a model for overall optimization of the classification schedule and sidings layout, simultaneously addressing the number of sorting steps and the total number of wagon movements in the objective function. The model is created on a binary integer programming formulation. A Variable Neighborhood Search (VNS) heuristic is developed due to the computational complexity of the problem. Numerical experiments validate the formulated model and demonstrate the efficiency of the VNS meta‐heuristic approach.