Modeling and Comparison of Different Management Approaches on the Intersections Network

The aim of the work is to develop a set of algorithms and programs for simulation modeling of traffic flows and traffic optimization for sections of the Krasnoyarsk road network. The paper investigates models of the process of selecting the phases of traffic lights in order to minimize the travel time of vehicles through the road network section. As a way to control traffic light objects, a phase selection process has been proposed based on the Multiagent Reinforcement Learning for Integrated Network also known as MARLIN. As an alternative to the traffic light phases change process in MARLIN, a fuzzy model of phase duration change on the next cycle is proposed. A microscopic Intelligent Driver Model (IDM) was used to simulate road traffic. To compare the models, a series of computational experiments were performed on the example of a section of the Krasnoyarsk road network.