Artificial Intelligence and Data Science
Online ISSN : 2435-9262
Development and validation of a mathematical optimization model for selecting rail replacement segments
Takafumi YOSHIDA, Daiki SAITO, Mami MATSUMOTO, Ryo YAMAGATA
Author information
JOURNAL OPEN ACCESS

2026 Volume 7 Issue 3 Pages 523-532

Details
Abstract

Railway rails are subject to defects caused by rolling contact fatigue with wheels, and their progression may lead to rail fracture. Therefore, railway operators conduct periodic inspections to identify the location and condition of defects and remove them through measures such as rail replacement. In this study, a math-ematical optimization model is developed to determine rail replacement segments for removing defects, considering the location and condition of rail defects in tracks with continuously welded rails. The model is formulated as a multi-objective optimization problem that minimizes the number of replacement seg-ments and the total replacement length, while maximizing the number of removed defects and the total severity of the removed defects. Furthermore, through numerical experiments and application to real rail-way data, it is demonstrated that the proposed method can generate maintenance plans that balance rail replacement workload and defect removal effectiveness. The results indicate that the proposed approach has potential as a decision-support tool for practical railway maintenance planning.

Content from these authors
© 2026 Japan Society of Civil Engineers
Previous article Next article
feedback
Top