Neuronal network approach to study the operation of shuttle-service trains
| Autor: | A. L. Kazakov, R. Yu. Upyr, A. D. Domojirova |
|---|---|
| Jazyk: | ruština |
| Rok vydání: | 2023 |
| Předmět: | |
| Zdroj: | Вестник Научно-исследовательского института железнодорожного транспорта, Vol 82, Iss 2, Pp 158-167 (2023) |
| Druh dokumentu: | article |
| ISSN: | 2223-9731 2713-2560 |
| DOI: | 10.21780/2223-9731-2023-82-2-158-167 |
| Popis: | Introduction. The authors study the section of the shuttle-service trains. Based on a detailed description and characterisation of the site, it is categorised as a man-machine and difficult-to-formalise system. It is not possible to study stability of such transport facilities using known criteria and methods. This study aimed at developing a new approach for assessing sustainability of a shuttle-service train system, taking account of its specific features and characteristics.Materials and methods. The paper proposes the use of a neural network approach to assess the sustainability of shuttle-service trains. The input data used is a time series of freight flow on a section of the shuttle-service trains, which is the main operational indicator of the system under study. The method of artificial neural networks is chosen for modelling because it is a simple and effective tool for studying the system in question, with the values of the freight flow having a random character.Results. An approach based on the power of neural networks is described. This indicator enables to assess quality of the model by reflecting the consistency of model results with actual data.Discussion and conclusion. The approach presented may be used to assess the sustainability of a shuttle-service train system. |
| Databáze: | Directory of Open Access Journals |
| Externí odkaz: |
|