Reconciliation of Train-Run data and Train consumption data for a better traction energy consumption estimation, a new Tool

Abstract

The present work addresses the estimation of the traction energy using train-run data. The proper quantification of electrical traction energy by Railway Undertaking is necessary for energy settlement and billing by the different energy stakeholders such as energy suppliers as well as the infrastructure managers. The system thus needs a fair estimation of the train consumption most notably in case of missing data. Train-run data are used to compute an estimation at the train level and then aggregated at the Railway Undertaking level. We used an aggregate tree-based model for the estimation models using a set of different physical variables extrapolated from a train dynamic physical model. In turn, it has imposed another tool for the reconciliation between train-run data and consumption data at the vehicle level because often train-run data and consumption data are siloed.This methodology has allowed the development of ORES, a new tool dedicated to the train-run data and consumption data reconciliation as well as train-run electrical traction energy estimation. Some additional approaches have also been tried using consumption at substations. The tools have been successfully used on various types of mission and rolling stocks and have enabled to better understand the limits of such systems.