№68-16

Research of dynamic signals of single platform railway scales

I. Kolysnychenko

1 Dnipro University of Technology,Dnipro, Ukraine

Coll.res.pap.nat.min.univ. 2022, 68:174-183

https://doi.org/10.33271/crpnmu/68.174

Full text (PDF)

ABSTRACT

Objective. Using experimental data obtained from existing systems installed at Ukrainian enterprises, it is necessary to conduct research on empirical data and obtain an algorithm for approximating the passage of wagons, thanks to which it is possible to obtain approximating functions for each type of bogie and automatic coupler used on the territory of the country and restore data on the passage of moving objects through a weighing platform with a minimum error for further use of the obtained results in building a system for identifying cars based on machine learning.

Research methods. Numerical methods are used to approximate data, obtain from the weighing system and determine various types of moving objects through single-platform railway scales in motion, namely, approximation to the experimental data of the Heaviside estimate.

The software developed in the Python programming language using the numPy library.

Findings. Using numerical methods, it was possible to obtain an algorithm for approximating the passage of automatic couplers for various combinations of automatic couplers and car bogies separately.

Thanks to the normalization of the data from the sensors over time, it was possible to avoid the dependence of the final results on the speed of the passage of the wagon, expressing as a percentage the dependence of the stay of the axles of the car on the weight platform, which made it possible to identify different types of cars with the same axle, but different characteristics (car base, bogie base), using the ratio of the time spent by the axes on the weighing platform.

The originality. The novelty lies in obtaining an algorithm for approximating experimental data on the passage of railway bogies and automatic couplers through single-platform scales, which can be used as a dataset generator close to real data on the passage of a railway train for their further use in setting up machine learning models.

Practical implications. Improving the accuracy of identifying railway moving objects, classifying them not only by the number of axles, but also by type, based on its unique overall characteristics, reduces weighing errors and downtime of the enterprise, which contributes to an increase in the number of weighed mobile railway objects.

Keywords: scales, weighing platform, railway car, Heaviside function, approximation, error estimation, identification, Python, dynamics.

References

  1. Sekuła, K., Wiącek, D., & Motylewski, J. (2018). In-motion rail scales as a component of the railway bridge diagnostic system. Scientific Journal of Silesian University of Technology. Series Transport, 101, 203–213.
    https://doi.org/10.20858/sjsutst.2018.101.18
  2. Yanik, A., & Higgins, C. (2019). Reliability analysis of oregon bridges using WIM (weigh-in-motion) data. El-Cezeri Fen ve Mühendislik Dergisi.
    https://doi.org/10.31202/ecjse.577400
  3. Hajializadeh, D., Žnidarič, A., Kalin, J., & OBrien, E. J. (2020). Development and testing of a railway bridge weigh-in-motion system. Applied Sciences, 10(14), 4708.
    https://doi.org/10.3390/app10144708
  4. Sposob vzveshivaniya zheleznodorozhnykh ob''ektov (2010). (Patent Rossiya № RU2390735C1).
  5. Zhang, W., Li, C.-l., Di, X.-f., Chen, M., & Tao, S. (2017). Research on automotive dynamic weighing method based on piezoelectric sensor. MATEC Web of Conferences, 139, 00203.
    https://doi.org/10.1051/matecconf/201713900203
  6. Lunys, O., Dailydka, S., Steišūnas, S., & Bureika, G. (2016). Analysis of freight wagon wheel failure detection in lithuanian railways. Procedia Engineering, 134, 64–71.
    https://doi.org/10.1016/j.proeng.2016.01.040
  7. Kolysnychenko, I., & Tkachov, V. (2021). A Polynomial approximation of dynamic signals of single platform railway scales. Electrical Engineering and Power Engineering, (2), 44–52.
    https://doi.org/10.15588/1607-6761-2021-2-5
  8. VanderPlas, J. (2016). Python data science handbook: Essential tools for working with data (D. Schanafelt, Ред.). O’Reilly Media.
  9. Mazurov, B. T., & Padve, V. A. (2017). Metod naimen'shikh kvadratov (statika, dinamika, modeli s utochnyaemoy strukturoy). Vesnik SGUGiT, 22(2), 22–35.