Railway Maintenance Management Using a Stochastic Geometrical Degradation Model

Abstract

Railway transportation has a significant advantage compared to other modes of transportation due to its high level of safety, significant capacity, and environmentally friendly nature, leading to an ever increasing expansion of passenger numbers. To maintain a high safety level and utilize the maximum capacity of this system, maintenance management of railroad is of crucial importance. The most costly part of railroad management is related to track maintenance, and the most crucial parameter of tracks is their geometry. The main objective of this research is to develop a railroad management system to both minimize track-maintenance costs and sustain high track geometrical quality. First, a track-quality database for a sufficient period of time is acquired. Having applied this database, a track degradation model is developed. Cost and improvement models associated with different track-maintenance actions are determined. Using degradation, cost, and improvement models, maintenance planning is conducted through optimum selection of tracks, appropriate time, and suitable maintenance actions in order to minimize cost and maximize the track geometrical quality. This selection is a combinatorial problem with a large number of feasible solutions, which cannot be solved using conventional algorithms. A heuristic method such as the genetic algorithm is proposed herein and utilized as a powerful tool to solve this problem.