Search

This paper introduces an implementation of a system allowing to denoise the intensity data sensed by a DAS system before analyzing it through a convolutional neural network for train tracking. The results obtained showed that it is possible to track a train by focusing only on the magnitude of signals without caring out complex treatments.

We demonstrate the reliability of two coherent detection DAS systems on detecting vibrations generated in a railway environment. Systems have been validated by accelerometers based measurements. A demonstration of a rail vehicle tracking is also performed.

This article introduces a two-step method for automatic spatio-temporal events localization, in railway environment based on Optical DAS system. A preprocessing is applied to remove unnecessary information. Then, the watershed algorithm is used to locate and delimit the events. This allows to filter the data to treat which could then be pushed to a machine learning system for events recognition.

An optical frequency modulation approach is adopted to reduce the impact of the phase- OTDR interference fading. This approach is validated by phase analysis results from train detection in a railway tunnel in Paris.

The electromagnetic shielding effect of rolling stock control cabs equipped with coated windows panes for heating purposes was investigated. Indeed such new generation windows exhibit higher levels of electromagnetic shielding leading to higher radio signals attenuations. Experimental data have been collected during measurement campaigns in order to characterize rolling stock cabs electromagnetic behavior. Measurement results show that an attenuation of the electromagnetic field can be assessed and is dependent on incident wave polarization. Those results represent a first step in electromagnetic behavior evaluation of train control cabs and will constitute a basis for future experiment and numerical simulations.

This paper deals with the electromagnetic (EM) disturbances which can affect the capability of the GSM-R communication system. It is based on a large number of data collected during measurement campaigns on board moving trains. On the one hand, the nature of disturbances received by the GSM-R antenna on board are classified and analyzed in order to extract a risk analysis concerning the noise level which can be produced on the GSM-R frequency channels. On the other hand, GSM-R coverage level measured on board train will also be presented and confronted with the noise level observed.