ADAPTIVE TRAFFIC LIGHT CONTROL ALGORITHM BASED ON THE DETECTION OF CARGO TRANSPORT

The application of computer vision and traffic flow modeling with an adaptive traffic light control algorithm for prioritizing the movement of cargo transport is described. The study objective is to increase the efficiency of traffic management in the cities of the oil and gas complex by developing a traffic light control algorithm that gives priority to cargo vehicles. Python, Ultralytics framework (YOLO), OpenCV, and libraries for working with data (Pandas, NumPy) are used to solve computer vision problems. For micromodeling of traffic flows, PTV Vissim is used together with VisVap or LISA+ module for adaptive traffic light control. With the introduction of priority passage for cargo vehicles at a controlled intersection, the traffic parameters for the general traffic flow change slightly, while for cargo vehicles the changes are significant. With an increase in the traffic intensity of cargo vehicles, priority travel has a positive effect on the traffic parameters of the entire traffic flow. With the use of neural network technologies and the integration of data into technical means of traffic management, it is possible to increase the average speed of traffic flows.