Разработка метода и определение нового условия безопасности движения в состоянии многопоточности на основе математического моделирования

Модели, системы, сети в экономике, технике, природе и обществе. 2023. № 3. С. 98–114.

Zykov S. V., Золотухина М. А., Золотухин С. А.

Background. All attention is focused on reducing the number of accidents on
an unfavorable section of the Altufyevsky highway in the area of 5 intersections. Such incidents
become aggravating circumstances for nearby infrastructure facilities, such as: shopping
malls, shops, cafes, there are also metro stations on a busy site. The endless flow of
people and cars, public transport stops, etc. require new security conditions. Materials
and methods. To carry out the analysis, methods of modeling, statistics and data mining were
used, namely: to simulate the traffic flow on a particular road section, to determine the problem
of the site, and to apply the simulation methodology to derive an algorithm for improving the
quality of traffic, defining new safety conditions and improving the characteristics of determining
the parameters of adverse events – these are the main aspects of the study. Results. A model
of multithreaded traffic in a software environment has been developed, an analysis of the traffic
flow has been carried out to improve the method used and to derive new conditions for safe traffic
on multithreaded road sections and intersections. A decision-making model is derived by creating
a training sample from the video stream of the characteristics presented in the article. Conclusions.
The presence of traffic lights is one of the safety conditions, but it has some nuances. All stages
of detection of adverse situations and new solutions are spelled out in the article.

Keywords: машинное обучение, имитационное моделирование, нейронная сеть, анализ данных, neural networks, simulation modeling, data analysis, machine learning, car accidents, управление безопасностью, safety management, автомобильная авария