Главная | Публикации | Development of a human flow generation module for testing machine learning algorithms

Development of a human flow generation module for testing machine learning algorithms

Авторы

Kulagin K., Leksashov A.V., Bolsunovskaya M.V., Shirokova S.V.
Transportation Research Procedia. 2021. Т. 54. С. 783–792.

Краткое описание

This article describes a custom software module for generating human flows that is currently under research and development. Existing works on this topic are investigated. The main software developments for generating human flows and modern development tools that can help in creating the necessary module are analyzed. The choice of the development environment, language, third-party programs and modules that help in development is justified. The developed customizable module of human flows provides the movement of people, including traffic across the road, the ability to select certain groups of people to place them on the street. In addition, this module has been integrated with the virtual urban environment and transport generation module to test connectivity with other projects. Functional and load testing carried out at the end revealed some shortcomings, which were later corrected. The authors make assumptions about the integration of the module into a complex traffic model.

Ключевые слова

Human flow generation module, integrated traffic system, software module, testing machine learning algorithms

Kulagin K., Leksashov A., Bolsunovskaya M., Shirokova S. Development of a Human Flow Generation Module for Testing Machine Learning Algorithms // Transportation Research Procedia. V. 54. 2021. P. 783-792. DOI:10.1016/j.trpro.2021.02.134.