On the Problem of Developing a Realistic Road Infrastructure Simulator for Reinforced Learning
Ksenia Polyantseva, Mikhail Gorodnichev, Yaroslav Mitrokhin, Marina Moseva
Abstract
One of the serious problems faced by urban municipalities is traffic congestion. It makes life in cities inconvenient for people. Every year, the government spends large budgets to solve this problem. Because of poorly planned road networks, a common result in many developing regions is small critical areas that are often hotspots of congestion; poor traffic management around these hotspots can lead to prolonged congestion. The work aims to develop a traffic flow modeling environment for the street and road network. This paper presents an analysis of already created solutions and tools for realistic urban simulation environments. The paper describes the stages of simulator design: requirements, methods of creation, functionality. The implementation of simulator components is described: development of the application and interaction subjects, software testing.