Improving the Decision-Making System of ITS Based on Dynamic Modeling of the Intensity of Vehicle Traffic along the Lanes, Taking into Account Environmental Factors

The problem of traffic congestion formation taking into account environmental factors in cities remains unresolved to the end. One of the tools of intelligent transport systems (ITS) for solving such problems is the dynamic management of traffic flows at urban controlled intersections. However, despite the existing experience in the use of ITS, there are currently no practical methods that take into account the influence of the concentration of traffic flow and emissions of harmful substances over time. The possibility of using neural networks to implement the control process is also insufficiently studied. Therefore, the purpose of this study is to develop a methodology for managing traffic flows at urban controlled intersections at different time concentrations. As a basis for obtaining a new technique, the Webster algorithm and obtaining big data on the parameters of traffic flows based on neural networks in real time were chosen. The developed technique has a number of distinctive features. The authors proposed to implement management according to the criterion of the optimal value of the band occupancy, taking into account the minimum impact on the quality of atmospheric air. The initial data is collected from the video stream from street surveillance cameras and the data is processed using neural network technologies. An analytical evaluation of the effectiveness of the research results was carried out. It is established that the application of the technique will enable positive technological, economic and environmental effects.