Concept of Digital Platform at Marshalling Yards

Implementation of the Industry 4.0 concept is considered in the context of automation of railway transport. The analysis refers to prerequisites for creation of a universal digital platform integrating automation systems at a marshalling yard.

The example of JSC Russian Railways has contributed to describe the main goals of Digital Station concept, aimed at fusion of data from low-level local automation equipment. The presented functionality of the system for control and processing information on movements of wagons and locomotives at the station in real time (SCPI MWL RT) implements the set goals by integrating initial data from all automation and centralised traffic control systems operating at the station, checking it for consistency, eliminating information redundancy and generating in real time the current model of a marshalling yard regarding trains and wagons and based on data «from the wheel».

Description of the existing functionality of SCPI MWL RT, implemented at a facility, is followed by the analysis of the advantages of this system for the railway cargo transportation network. The objective of the paper is to present some previously unpublished technical solutions for implementation of the specified functionality. The methods of the research are based on fusion of heterogeneous data received from floor devices, specialised video cameras, as well as from real-time wagon positioning models.

It is shown that adoption of new technical solutions for SCPI MWL RT will allow to considerably improve the quality of planning of technological process of classifying railway wagons and of forecasting the need for infrastructure maintenance. Deep learning algorithms presented ensure functioning of the developed solutions in real time with high accuracy. Further steps described refer to implementation of a digital platform in the form of a digital twin of a marshalling yard, creating thus a prerequisite for development of an intelligent automatic machine to control the marshalling yard, as well as for further planned ways to implementation there-of.

1. Shabelnikov, A. N., Sukhanov, A. V. Components of cyber-physicalsystems as part of KSAU SP[Komponenty kiberfizicheskikh system v sostave KSAU SP]. Avtomatika, svyaz’, informatika, 2020,Iss. 1, pp. 12–14. DOI: 10.34649/AT.2020.1.1.002.

2. Rozenberg, I. N., Shabelnikov, A. N. Digital marshalling yard [Tsifrovaya sortirovochnaya stantsiya]. Zheleznodorozhniy transport, 2018, Iss. 10, pp. 13–17.

3. Rozenberg, E. N., Batraev, V. V. Development of promising control systems and ensuring safety of train traffic [Razrabotka perspektivnykh system upravleniya i obespecheniya bezopasnosti dvizheniya poezdov]. Bulletin of Joint Scientific Council of JSC Russian Railways, 2017, Iss. 4, pp. 43–51. [Electronic resource]: https://elibrary.ru/download/elibrary_30763584_22331826.pdf. Last accessed 29.05.2020.

4. Shabelnikov, A. N., Sokolov, V. N. KSAU SP – a newdirection of automation of humps[KSAU SP – novoe napravlenie avtomatizatsii sortirovochnykh gorok]. Avtomatika, svyaz, informatika, 2017, Iss. 8, pp. 2–4.

5. Zamyshlyaev,A. M., Kalinin,A. V., Dolganyuk, S. I. MALS system: problems and perspectives[Sistema MALS: zadachi i perspektivy]. Avtomatika, svyaz, informatika, 2016, Iss.10, pp. 30–33.

6. Anoshkin, V. V., Shabelnikov, A. N., Shipulin, N. P. Interactive hump console [Interaktivniy gorochniy pul’t]. Avtomatika, svyaz, informatika, 2020, Iss. 5, pp. 11–12.

7. Shabelnikov, A. N., Smorodin, A. N., Complex automation of a junction marshalling yard [Kompleksnaya avtomatizatsiya uzlovoi sortirovochnoi stantsii]. Avtomatika, svyaz, informatika, 2018, Iss. 4, pp. 12–14.

8. Shabelnikov, A. N., Olgeizer, I. A., Rogov, S. A. From mechanisation to digitalisation of a marshalling yard [Ot mekhanizatsii k tsifrovizatsii sortirovochnoi stantsii]. Avtomatika, svyaz, informatika, 2018, Iss. 1, pp. 21–23.

9. Ronneberger, O., Fischer, P., Brox, T. U-Net: Convolutional Networks for Biomedical Image Segmentation. International Conference on Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015.MICCAI 2015. Lecture Notesin Computer Science, Springer, Cham, 2015, Vol. 9351, рр. 234–241. [Electronic resource]: https://link.springer.com/content/pdf/10.1007%2F978-3-319-24574-4_28.pdf. Last accessed 29.05.2020.

10. Sergeev, A., Del Balso, M. Horovod: fast and easy distributed deep learning in TensorFlow // arXiv preprint arXiv:1802.05799, 2018. [Electroic resource]: https://arxiv.org/abs/1802.05799. Last accessed 29.05.2020.

11. Olgeizer, I. A. Digital twin of a sorting hump [Tsifrovoi dvoinik sortirovochnoi gorki]. Avtomatika, svyaz, informatika, 2020, Iss. 1, pp. 20–22.