Fuzzy logic-based controller of traffic intersection

Abstract

The widely used conventional control of traffic intersections is based on classic logic. The strategy of this approach can be divided in two main procedures: fired-time system and on-line system. The fired-time system uses predefined time intervals to control the car-flow through the intersection. The on-line system utilises the proximity sensors and combines the predefined time intervals with changes in the particular cycles. Usually, sets of measurements must be taken to determine the proper adjustment of time cycles. The measurements are based an observations of flow patterns and car counting during different periods of the day. The major inconvenience of the current traffic lights control is the low accuracy, allowing only crude changes in the green-red cycle. Fuzzy logic control can be used as an alternative approach to the traffic environment. A fuzzy controller can provide smoother and more flexible control of the timing of the green-red phases, depending on the flow density of vehicles. The purpose of the proposed method is to minimize the waiting time of vehicles in an intersection.