The Use of Neural Network Method in Determining the Proportion of Traffic Flow Redirection in the Road Network

The article presents the results of an experimental study on traffic management in the event of unpredictable situations in the transportation network. The aim of this research is to determine the proportion of traffic flow through the network in the situation of reduced capacity of road segments (UDS), including due to traffic accidents. This work proposes a method to address congestion issues by redirecting vehicles from roads with high network load coefficients (z) to alternative routes, while determining the volume of traffic flow to eliminate the possibility of congestion in other transportation network segments. The method of survey and regression analysis using neural networks is employed to determine the proportion of traffic flow redirection. The article examines the possibility of addressing congestion in the city's transportation network by improving the efficiency of indirect traffic management systems, which are part of the Intelligent Transportation System (ITS). The proposed method will enhance the efficiency of ITS by rerouting vehicle movement in conditions of high risk of congestion and traffic accidents.