Grey-system-theory-based model for the prediction of track geometry quality

The quality of track geometry is an important aspect in railway engineering, as it reflects any deviations and thus the actual condition of a track. Monitoring and prediction of a relevant geometry quality parameter provides an opportunity for effective maintenance, thus creating the advantages of extending the life of the asset, reducing maintenance costs and minimizing possession time requirements. Effective maintenance practice requires a good understanding of the behaviour of track structures over time and also prediction of its condition using only a few inputs. This paper presents a grey-system-theory-based model for predicting track irregularity. Three variants of the grey model are presented and their performances are compared with simple linear and exponential models. Regression models and the grey-system-theory-based models are used to obtain the standard deviation of the longitudinal level from a series of geometry inspection data. The overall performances of the models are evaluated in terms of the regression and prediction accuracies, and it is shown that a Fourier series modification of the grey model has the best performance and the minimum error. The contribution of this paper is the creation of a prediction model for track geometry quality, which is essential for planning and scheduling of preventive geometry maintenance.