Article

Article title INTELLIGENT APPROACH FOR FAULT PREDICTION IN THE PROCESS OF BREAKING-UP THE TRAINS AT HUMP YARDS
Authors A.N. Shabelnikov, S.M. Kovalev, A.V. Sukhanov
Section SECTION III. SAFETY OF COMPLEX SYSTEMS
Month, Year 08, 2016 @en
Index UDC 656.212.5
DOI 10.18522/2311-3103-2016-8-6878
Abstract Nowadays, a wide implementation of automation tools into the railway transportation sys-tems occurs. One of the important sides of this process is automation of the control for and trains makeup-breaking-up at hump yards in real-time mode with utilization of modern intelligent tech-nologies. “Intelligence incorporation” for technological processes, firstly, deals with the problem of decision-making support in complicated technological situations and, particularly, in abnormal situations, which referred to the faults in the process behavior. Here, the most important task is the immediate detection of such situations, which aim is to prevent faults and adjust the process. The paper proposes a new approach for preventive detection of abnormal situations in the behavior of the technological process of train breaking-up at hump yards, when the faults are related to over-taking of cuts. The proposed approach is based on the change of the method of fault prediction by the method of precursor detection. The detection method is based on granular model of process representation in form of specific temporal patterns. Granular model of technological process is formed on the basis of fuzzy estimates inference and fuzzy estimates postprocessing by using the tools of pragmatic logic. Precursors detection for the granular model is performed by using the model of multistep prediction. Here, multistep prediction is performed by utilization of Markov reward models tuned by temporal-difference learning algorithm. Computational experiments show that presented approach allows improving the efficacy of the decision making at hump yards. Moreover, it should be noted that proposed models and algorithms are usable for many other tasks, where decision making in uncertain conditions is required.

Download PDF

Keywords Breaking-up process; overtaking of cars; granular model; cognitive measurements; anoma-lous temporal pattern; stochastic model; temporal model.
References 1. Razrabotka planovykh zadaniy dlya zheleznodorozhnoy sortirovochnoy stantsii [Development of scheduled tasks for the rail yard], TRGREAT. Available at: trgreat.ru/gov-1199.html (accessed 10 September 2015).
2. Shirokova V.V., Nesvetova E.A. Organizatsiya raboty sortirovochnoy stantsii [Organization of work the rail yard]. Khabarovsk: Izd-vo DVGUPS, 2006, 98 p.
3. Kondrat'eva L.A. Ustroystva zheleznodorozhnoy avtomatiki i telemekhaniki. (Obshchiy kurs): uchebnik dlya tekhnikumov zh.-d. transp. [Devices of railway automatics and telemechanics. (General course): textbook for technical schools of railway transport]. Moscow: Transport, 1983, 232 p.
4. Tulupov L.P., Letskiy E.K., Shapkin I.N., Samokhvalov A.I. Upravlenie i informatsionnye tekhnologii na zheleznodorozhnom transporte [Management and information technology in railway transport], ed. by Tulupova L.P. Moscow: Marshrut, 2005, 467 p.
5. Kovalev V.I., Os'minin A.T., Groshev G.M. Sistemy avtomatizatsii i informatsionnye tekhnologii upravleniya perevozkami na zheleznodorozhnom transporte [Automation systems and information technology management in railway transportation]. Moscow: Marshrut, 2006, 544 p.
6. Ivanchenko V.N., Kovalev S.M., Shabel'nikov A.N. Novye informatsionnye tekhnologii: integ-rirovannaya informatsionno-upravlyayushchaya sistema avtomatizatsii protsessa rasformiro-vaniya–formirovaniya poezdov: uchebnik [New information technologies: integrated information management system automate the process of dissolution–formation: textbook]. Rostov-on-Don: RGUPS, 2002, 276 p.
7. Shabel'nikov A.N., Sukhanov A.V., Kovalev S.M. Intellektual'nyy metod predskazaniya poyavleniya neshtatnykh situatsiy v protsesse rasformirovaniya poezdov na sortirovochnoy gorke [Intelligent method of predicting the occurrence of abnormal situations in the process of dissolution of trains on the hump], Inzhenernyy vestnik Dona [Engineering journal of Don], 2015, No. 4. Available at: http://ivdon.ru/ru/magazine/archive/n4y2015/3334 (accessed
12 November 2015).
8. Chervotenko E.E. i dr. Proektirovanie sortirovochnykh ustroystv: ucheb. Posobie [Design of screening devices: a tutorial]. Khabarovsk: Izd-vo DVGUPS, 2014, 75 p.
9. Shabel'nikov A.N. i dr. Sistemy avtomatizatsii sortirovochnykh gorok na osnove sovremennykh komp'yuternykh tekhnologiy: uchebnik dlya vuzov zh.-d. transporta [Systems of automation of hump yards on the basis of modern computer technology: textbook for universities of railway transport]. Rostov-on-Don: NIIAS. Rost. gos. un-t putey soobshcheniya, 2010.
10. Ivanchenko V.N., Shabel'nikov A.N. Novyy podkhod k postroeniyu intellektual'nykh informatsionno-upravlyayushchikh sistem na zheleznodorozhnom transporte [A new approach to building intelligent information management systems in railway transport], Izvestiya vysshikh uchebnykh zavedeniy. Severo-Kavkazskiy region. Tekhnicheskie nauki [University News. North-caucasian region. Technical Sciences Series]. Appendix No. 2, 2004, pp. 109-116.
11. Lominoga, I.V. Algoritm rascheta ekonomicheskikh poter' na sortirovochnoy gorke [The algo-rithm of calculation of economic losses on the hump], Problemy sovremennoy ekonomiki: Ma-terialy III Mezhdunar. nauch. konf. [problems of modern Economics: materials of the III Mezhdunar. scientific conference]. Chelyabinsk: Dva komsomol'tsa, 2013, pp. 102-105.
12. Ivanchenko V.N. Sovremennye informatsionnye tekhnologii upravleniya slozhnymi protsessami rasformirovaniya-formirovaniya poezdov [Modern information technologies in the management of complex processes of dissolution-formation], Nauka i transport. Modernizatsiya zheleznodorozhnogo transporta [Science and transportation. The modernisation of railway transport], 2013, No. 2 (6), pp. 64-69.
13. Kovalev S.M. Shabel'nikov A.N. Teoreticheskie problemy intellektualizatsii transportnykh protsessov [Theoretical problems of intellectualization of transport processes], Avtomatizatsiya i mekhanizatsiya tekhnologicheskikh protsessov na sortirovochnykh stantsiyakh: Trudy Mezhdunarodnoy nauchno-prakticheskoy konferentsii [Automation and mechanization of technological processes in marshalling yards: proceedings of the International scientific-practical conference]. Moscow, 2010, pp. 15-19.
14. Shabel'nikov A.N., Sukhanov A.V., Kovalev S.M. Intellektual'nyy metod predskazaniya poyavleniya neshtatnykh situatsiy v protsesse rasformirovaniya poezdov na sortirovochnoy gorke [Intelligent method of predicting the occurrence of abnormal situations in the process of dissolution of trains on the hump], Inzhenernyy vestnik Dona [Engineering journal of Don], 2015, No. 4. Available at: http://ivdon.ru/ru/magazine/archive/n4y2015/3334 (accessed
12 November 2015).
15. Kovalev S.M. Metody mnogoshagovogo predskazaniya anomaliy v temporal'nykh dannykh [The method of multistep prediction of anomalies in temporal data], Izvestiya YuFU. Tekhnicheskie nauki [Izvestiya SFedU. Engineering Sciences], 2013, No. 7 (144), pp. 85-91.
16. Yeung D.Y., Ding Y.X. Host-based intrusion detection using dynamic and static behavioral models, Pattern Recognition, 2003, No. 36, pp. 229–243.
17. Kovalev S.M., Sukhanov A.V. Obnaruzhenie osobykh tipov patternov vo vremennykh ryadakh na osnove gibridnoy stokhasticheskoy modeli [Special temporal pattern recognition technique based on hybrid stochastic model], Izvestiya YuFU. Tekhnicheskie nauki [Izvestiya SFedU. Engineering Sciences], 2014, No. 4 (153), pp. 142-149.
18. Xu X. Sequential anomaly detection based on temporal-difference learning: Principles, models and case studies, Applied Soft Computing, 2010, Vol. 10, No. 3, pp. 859-867.
19. Sukhanov A.V., Kovalev S.M., Styskala V. Advanced Temporal-Difference Learning for Intrusion Detection, 13th IFAC and IEEE Conference on Programmable Devices and Embedded Systems PDES. IFAC-PapersOnLine, Elsevier, 2015, No. 48 (4), pp. 43-48.
20. Kovalev S.M., Sukhanov A.V. Anomaly detection based on Markov chain model with production rules, Программные продукты и системы [Software Products and Systems], 2014,
No. 3, pp. 40-44.

Comments are closed.