Application of Artificial Neural Networks in Predicting the Degradation of Tram Tracks Using Maintenance Data

Sara Moridpour

RMIT University, Australia

Ehsan Mazloumi

RMIT University, Australia

Reyhaneh Hesami

Yarra Trams, Australia

ABSTRACT

The increase in number of passengers and tramcars will wear down existing rail structures faster. This is forcing the rail infrastructure asset owners to incorporate asset management strategies to reduce total operating cost of maintenance whilst improving safety and performance. Analysing track geometry defects is critical to plan a proactive maintenance strategy in short and long term. Repairing and maintaining the correctly selected tram tracks can effectively reduce the cost of maintenance operations. The main contribution of this chapter is to explore the factors influencing the degradation of tram tracks (light rail tracks) using existing geometric data, inspection data, load data and repair data. This chapter also presents an Artificial Neural Networks (ANN) model to predict the degradation of tram tracks. Predicting the degradation of tram tracks will assist in understanding the maintenance needs of tram system and reduce the operating costs of the system.

DOI: 10.4018/978-1-5225-0886-1.ch002

 
Source
< Prev   CONTENTS   Source   Next >