Predictive Maintenance, Prognostics and Reliability Centered Maintenance approach are becoming more and more important in the railway sector to reduce costs of operation and increase reliability and safety. In fact they are fundamental to optimize the maintenance process, defining new measures and algorithms to locate faults, monitor health conditions of subsystems and estimate residual life of components. However there is a tradeoff to identify: new measures and new algorithms imply new sensors and new processing devices, and these items have their own cost and their own reliability; this issue has to be taken into account to evaluate the global benefit. In some cases, however, it's possible to use existing sensors and existing processing hardware to extract new information from already available data. It's clear that this is usually the best option because the benefit can be achieved with little or not cost at all. This paper describes the result of a study performed with the aim of detecting arcing events without the need of additional equipment mounted on board the train. A set of data relative to voltage and current collected by trains in high-speed lines together with a set of measurements coming from photosensors are available. The data are processed by the use of an advanced classification technique, namely Support Vector Machines, with the aim of extracting important information such as the time coordinate related to anomalies in the overhead contact line and the status of the contact strip of the pantograph.

No abstract

Abstract-This work presents a channel model for the broadband characterization of power lines in presence of time variation of the loads. The model is characterized by taking into account both measured and geometrical channel characteristics and can easily be used to take into account also the presence of noise. The channel is described by a two-port equivalent described by a scattering matrix determined from a wavelet-based expansion of the input and output quantities. Upper and lower bounds for the response of the channel in presence of time-varying loads are determined in a fast and efficient way avoiding time consuming Monte Carlo simulations. The bounds determination allows the estimate of noteworthy quantities for the tuning of currently used modulation schemes for power lines communications such as orthogonal frequency-division multiplexing.

No abstract

A crucial problem for railway companies is the detection of the critical points of the overhead line for predicting excessive wear of the contact wire and for helping maintenance activities. An innovative tool, based on wavelet multiresolution analysis, is proposed in this paper for on-line monitoring of the quality of current collection. Wavelet decomposition of data collected from different measurement systems (e.g. total traction line current or signals acquired from unconventional sensors) can perform an automatic analysis of the data collected and can detect a singular occurrence of arcing due to loss of contact of the pantograph. A relevant outcome of such an analysis is the evaluation of a quantitative index for checking the quality of the current collection via an integral analysis of the multiresolution signal decomposition. Examples of analysis from experimental data are reported in the paper.