Continuous improvements in transport systems suggest the need for monitoring, evaluation of traffic congestion, regular traffic jams during peak load hours, low traffic potential. In this regard, the formation and development of intelligent transport systems based on the training of neural networks in the objective assessment of the parameters of the operation of transport systems within optimally established limits is a priority in the study of transport flows. The purpose of the chapter is the research to solving problems of creating intelligent systems based on training neural networks. There is the analysis of the current trends in the modeling of transport systems in the chapter. A neural network of the transport system is built. The research materials can be used in the development of software complexes of intelligent transport systems that allow to track the dynamics of traffic congestion, to determine the cause-and-effect relationships of deviations from the classical theory, to establish the optimal vector of stable traffic and the parameters of reserves for increasing traffic potential.