Adaptive traffic light control on highway entrances

In this work, we study the problem of optimal traffic signal control of automotive flow on urban multilane road. We employ macroscopic approach, describing traffic flow with non-linear second-order hyperbolic system on PDEs (for traffic density and speed). We close this system with the equation for dependency of traffic flow intensity on its density, obtained empirically for each segment of the road based on traffic detector data and GPS tracks. We verify the effectiveness of proposed methodology using numerical modeling of Moscow Automotive Ring Road, a 10-lane, 108 km long highway encircling the city of Moscow, Russia.