Predictive maintenance of railway catenary: the role of physical modeling

CEIT HAS DEVELOPED PHYSICAL MODELS OF CATENARY THAT ALLOW EVALUATING THEIR STATUS IN REAL-TIME USING SENSORIZED PANTOGRAPHS.

Predictive maintenance is involving a qualitative leap in infrastructure management. In this field, Ceit provides innovative solutions with high added value, relying on its knowledge of data analysis and advanced modelling of railway infrastructure and vehicles. Within the framework of the European project SIA (System for vehicle infrastructure Interaction Assets health status monitoring), Ceit has developed physical models of catenary that allow evaluating their status in real-time using sensorized pantographs. Specifically, work has been done on monitoring the stagger, height and wear of the contact wire, and the quality of the pantograph-catenary interaction.

However, predictive maintenance is not only intended to assess the state of assets but also to predict their evolution. One of the main challenges in its implementation is obtaining and managing large amounts of data and measurements, a task that can take several years before starting to make future predictions. At this point, physical models can be used to speed up and lighten this process by assisting in the creation of digital twins.

A digital twin based on measurements and physical models allows a great variety of future cases to be considered. Therefore, it will improve its ability to anticipate by using fewer measurements, and in turn, it will reduce the time in which forward-looking predictions become operational. The models developed by Ceit carry out these simulations by faithfully representing the Spanish catenary infrastructure, which largely uses double contact wire, and by introducing algorithms capable of greatly reducing simulation times. The ultimate goal is that these models combine precision, versatility and efficiency.