Combination of Data-driven Feature Selection Methods with Domain Knowledge for Diagnosis of Railway Vehicles

##plugins.themes.bootstrap3.article.main##

##plugins.themes.bootstrap3.article.sidebar##

Published Oct 2, 2017
Bernhard Girstmair Andreas Haigermoser Justinian Rosca

Abstract

Railway vehicles are generally maintained preventively within certain time periods. Condition based predictive maintenance strategies have a great economic potential so that modern trains are equipped with many sensors in order to perform diagnostics and prognostics of components.
Methods for fault detection need appropriate feature subsets in order to achieve small in-sample and out-sample errors. In our case the typical feature selection approach using pure data-driven methods is difficult, as the number of possible feature sets is very large. On the other hand there exists rich domain knowledge and detailed physical models of the mechanical system. The aim is to combine this knowledge with the often used mathematical methods for feature selection for improving classification of cases when a faulty damper is present. Based on the dynamic equations of motion, this paper presents heuristic feature selection via the analysis of transfer functions. We describe several wellknown methods of automated feature selection and a workflow which combines domain knowledge with automated methods. Results show that it is difficult to define
features based only on domain-knowledge, but in combination with data-driven techniques good classification performance can be achieved.

How to Cite

Girstmair, B., Haigermoser, A., & Rosca, J. (2017). Combination of Data-driven Feature Selection Methods with Domain Knowledge for Diagnosis of Railway Vehicles. Annual Conference of the PHM Society, 9(1). https://doi.org/10.36001/phmconf.2017.v9i1.2384
Abstract 258 | PDF Downloads 196

##plugins.themes.bootstrap3.article.details##

Keywords

feature selection, railway systems

References
Baumgartner J.P. (2001). Prices and costs in the railway sector, Ecole Polytechnique Federale de Lausanne.
Ellermann (2014). Mehrkörperdynamik, lecture script, Version 18, Graz University of Technology.
Fisher (1958). R.A. Statistical Methods for Research Workers, 13th Ed., Edinburgh : Oliver and Boyd.
Guyon, Bitter, Ahmed, Brown, and Heller (2003). Multivariate Non-Linear Feature Selection with Kernel Multiplicative Updates and Gram-Schmidt Relief proceedings of the BISC FLINT-CIBI 2003 workshop, Berkeley, Dec. 2003.
Guyon I., and Elisseeff A. (2003). An Introduction to Variable and Feature Selection, Journal of Machine Learning Research 3, pp. 1157-1182.
Guyon I., Weston J., Barnhill St., Vapnik V. (2002). Gene Selection for Cancer Classification using Support Vector Machines. Machine Learning Volume 46, Issue 1, pp. 389-422.
International Standards Organization (ISO) (2012). Condition Monitoring and Diagnostics of Machines -Prognostics part 1: General Guidelines. In ISO, ISO13379-1:2012(E). vol. ISO/IEC Directives Part 2, I. O. f. S. (ISO), (p. 2). Genève, Switzerland: International Standards Organization.
Iwnicki S. (2006). Simulation. In Polach O., Berg M. and Iwnicki S. Handbook of railway vehicle dynamics (pp. 359-423). Boca Raton London New York.
Kimothol J.K., and Sextro W. (2014). An approach for feature extraction and selection from non-trending data for machinery prognosis, Annual Conference of the Prognostics and Health Management Society. Sept 29 – Oct 02, Fort Worth.
Knothe, K., and Stichel S. (2003). Schienenfahrzeugdynamik. Berlin Heidelberg New York.
Massey, F.J. (1951). The Kolmogorov-Smirnov Test for Goodness of Fit, Journal of the American Statistical Association, Vol. 46, No. 253, pp. 68-78.
Peng H, Long F., and Ding C. (2005). Feature selection based on mutual in-formation: criteria of max-dependency, max-relevance, and min-redundancy, IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 27, No. 8, pp.1226-1238.
Schölkopf B., Williamson R., Smola A., Shawe-Taylor J. and Platt J. (1999). Support vector method for novelty detection. Proceedings of the 12th International Conference on Neural Information Processing Systems pp. 582-588. Nov 29 - Dec 04, 1999 Denver, CO.
Yan K., Zhang D. (2015). Feature selection and analysis on correlated gas sensor data with re-cursive feature elimination, Sensors and Actuators B: Chemical, pp. 353-363.
Yang W., Wang K. and Zuo W. (2012). Neighborhood Component Feature Selection for High-Dimensional Data, Journal of computers, Vol. 7, No. 1.
Section
Technical Research Papers