2021, Vol. 8, No. 3. - go to content...

Permanent address of this page - https://t-s.today/en/01sats321.html

Метаданные этой статьи так же доступны на русском языке

DOI: 10.15862/01SATS321 (https://doi.org/10.15862/01SATS321)

Full article in PDF format (file size: 1.4 MB)

For citation:

Polyanskiy A.V. Resource planning of the railway facilities construction technological process with the use of an artificial neural network. Russian Journal of Transport Engineering. 2021; 8(3). Available at: https://t-s.today/PDF/01SATS321.pdf (in Russian). DOI: 10.15862/01SATS321

Resource planning of the railway facilities construction technological process with the use of an artificial neural network

Aleksey V. Polyanskiy

Russian University of Transport, Moscow, Russia

Corresponding author: Aleksey V. Polyanskiy, e-mail: polal_82@mail.ru

Abstract. The article is dedicated to theoretical and practical research in the resource planning field for the railway facilities’ construction technological process using artificial neural networks. The study is a part of the subsystem in the development of railway construction engineering and technical maintenance — engineering and technical maintenance of railway facilities construction technological process. The subsystem is based on the effective automated systems use with elements of artificial intelligence. This is caused by deviations occurrence from the target requirements during the technological process implementation, due to the railway construction stochasticity, and the need for a prompt revision of the already made decisions. The existing methods allow us to correct the construction work organization, however, the technology remains unchanged, which is dictated by the design documentation and the work safety requirements. To give technological process flexibility in order to adapt to work changing conditions, it is necessary to provide an operational solution to the resource planning problem. The existing resource planning methods peculiarities, the current problem dimension, and the need to take into account a number of restrictions allow us to use artificial intelligence tools. In this regard, a methodology and a railway facilities construction technological process resource planning (labor, technical) computational-logical model with the artificial neural network use were developed. This approach is based on the lack of precise algorithmic actions or rules that can provide the desired result without formal complications. Also, the author has developed a special software module, prepared a data sample, an artificial neural network topological structure, and implemented an algorithm for its training, configuration, and testing. Based on the theoretical research and software module application results, the article presents the practical aspects of resource planning for the railway roadbed flooded embankment erection technological process.

Keywords: railway construction; technological process; resources; railway track object; nomenclature of works; artificial intelligence methods; artificial neural network

Download article in PDF format