[Back]

M. Tiefenbacher, M. Kozek:
"A Novel Approach to Extract Reference Trajectories from Measured Data Sets of Urban Light-Rail Systems";
Talk: 10th International Conference on Computational Structures Theory (CST 2010), Valencia (invited); 2010-09-14 - 2010-09-17; in: "Proceedings of the 10th International Conference on Computational Structures Technology", B.H.V. Topping, J.M. Adam, F.J. Pallarés, R. Bru, M.L. Romero (ed.); Civil-Comp Press, (2010), ISBN: 978-1-905088-37-9; Paper ID 41, 16 pages.

This paper introduces a new approach to retrieve both reference track geometry data and track irregularities from measured track data of urban light-rail systems. Urban light-rail systems are constructed under the limitations of historical topology of roads and other traffic zones. As a result the measured track cannot be classified into standardised geometric forms (straight lines, clothoids, curve segments) used for heavy rail systems. The presented framework provides methods to estimate the track irregularities and therefore enables track quality investigations. Furthermore, the reference
track geometry and complementary track irregularities are used to identify black-box models for lateral accelerations. Due to a non-linear input transformation of the reference track geometry and irregularities, the results of models with linear system dynamics
show comparable results to non-linear models identified with artificial neural networks.

railway engineering, track geometry, system identification, dynamic systems, neural networks, simulations.

http://publik.tuwien.ac.at/files/PubDat_193771.pdf