A Self-Adjusting Urban Traffic Control

Controlling traffic in cities is a very complex task, since efficient traffic flows require coordination of service requests from many participants. Traditional methods optimize signal timing plans, to maximize efficiency of traffic, based on predefined traffic flows e.g. fixed-time signal timing plans. Disadvantages of these methods are that they cannot cope with constantly changing traffic flows. Therefore, fixed-time signal timing plans are not robust enough to provide near-to-optimal service to fluctuating traffic flows. We propose a simple and feasible method which adjusts traffic lights to better utilize phase green splits. This method is based on detection of traffic flows and represents a form of actuated traffic control with a flexible phase sequence. The initial results show that self-adjusting traffic control significantly reduces average vehicle delays when compared with conventional traffic control. The method also shows some promising potential for coordinated traffic operations on urban arterials.