TY - JOUR TI - Multi-objective Optimization of Train Speed Profiles Using History Measurements AU - Achilleos, A. AU - Anastasopoulos, M. AU - Tzanakaki, A. AU - Iordache, M. AU - Langlois, O. AU - Pheulpin, J.-F. AU - Simeonidou, D. JO - Communications in Computer and Information Science PY - 2021 VL - 1217 TODO - null SP - 363-374 PB - Springer Science and Business Media Deutschland GmbH SN - null TODO - 10.1007/978-3-030-68028-2_17 TODO - Data envelopment analysis; Energy utilization; Green computing; Intelligent systems; Intelligent vehicle highway systems; Railroad transportation; Smart city; Traffic control, Model framework; Multi-objective optimization scheme; Nonparametric methods; Numerical results; Railway system; Scheduling constraints; Train speed, Multiobjective optimization TODO - The present study focuses on the development of a multi-objective optimization scheme to improve the efficiency of railway systems. This is achieved through the identification of the optimal train speed profiles employing a novel modeling framework based on Data Envelopment Analysis (DEA). Train speed profiles are selected with the objective to transfer more passengers in less time and with less energy under scheduling constraints. Given that DEA is a data oriented, non-parametric method, a large-scale experimental camping has been carried out over an operational tramway system to collect the required inputs/outputs. Numerical results show that when the proposed approach energy consumption can be reduced by 10%. © 2021, Springer Nature Switzerland AG. ER -