2026, Vol. 13, No. 2. - go to content...

Permanent address of this page - https://t-s.today/en/24sats226.html

Метаданные этой статьи так же доступны на русском языке

DOI: 10.15862/24SATS226 (https://doi.org/10.15862/24SATS226)

Full article in PDF format (file size: 2 MB)

For citation:

Polyanskiy A.V. Multi‑criteria selection of technological solutions for the construction of a railway subgrade using a genetic algorithm. Russian Journal of Transport Engineering. 2026; 13(2). Available at: https://t-s.today/PDF/24SATS226.pdf (in Russian). DOI: 10.15862/24SATS226

Multi‑criteria selection of technological solutions for the construction of a railway subgrade using a genetic algorithm

Polyanskiy Aleksey Viktorovich
Russian University of Transport, Moscow, Russia
E-mail: polal_82@mail.ru
ORCID: https://orcid.org/0000-0001-6190-0481
RSCI: https://elibrary.ru/author_profile.asp?id=412433
WoS: https://www.webofscience.com/wos/author/rid/AFJ-6625-2022
SCOPUS: https://www.scopus.com/authid/detail.url?authorId=57459964000

Abstract. The presented article is devoted to solving the scientific and practical problem of multi‑criteria selection of technological solutions for the construction of railway embankment on sections characterized by the presence of weak soils and requiring geosynthetic reinforcement. The relevance of the study is due to the high capital intensity of infrastructure projects, stringent regulatory requirements, and the need to take into account the nonlinear relationships between calendar deadlines, cost, labour input, and the maximum load on material and technical resources. The paper highlights the shortcomings of classical mathematical programming methods when solving problems with mixed discrete and continuous variables. A comprehensive mathematical model has been developed, which includes a formalized system consisting of four objective functions and strict constraints, describing the process of distributing five alternative technological solutions across twenty‑eight production sites of railway subgrade. Due to the fact that the formulated problem belongs to the class of NP‑hard combinatorial problems, a modified genetic algorithm for multi‑criteria optimization was used as the main computational tool. The text provides a detailed description of the algorithmic solution scheme, the software implementation of which was carried out in the Matlab environment with support for automated data exchange. The results of the conducted computational experiment demonstrate the successful construction of an approximation of the Pareto front, consisting of a set of non‑dominated plans. An analysis of engineering trade‑offs and the statistics of the use of technological solutions was carried out, which made it possible to identify objective patterns in their assignment depending on the volume of earthworks and engineering and geological conditions. Special attention was paid to the issues of verifying the obtained data and adapting the model to the real conditions of railway construction. The practical significance of the study lies in the development of a scientifically grounded methodological approach that provides a tool for making management decisions under conditions of uncertainty and conflicting target indicators.

Keywords: technological solution; technological process; railway construction; railway subgrade; artificial intelligence; multi‑criteria optimization; genetic algorithm

Download article in PDF format