Decentralised Car Traffic Control using Message Propagation and Re-routing

This paper describes a decentralised car traffic control simulation based on re-routing of vehicles and local propagation of messages among traffic nodes (roads intersections and traffic lights). A genetic algorithm is used to identify parameters´ values governing the simulations. This paper reports on the one hand on convergence results of the genetic algorithm, and on the other hand on preliminary comparison results of the fittest solutions.