Design of fuzzy logic based adaptive traffic signal controller

Murali Mohan Ande, University of Nevada, Las Vegas


Traffic control of street intersections is one of the most critical elements in providing an efficient flow of traffic in urban networks. Conventionally, pretimed controllers are used, but they cannot respond to real time fluctuations in traffic demand. Traffic actuated signals provide an improvement over pretimed controllers, but their performance deteriorates under heavy traffic conditions. These conditions necessitate the development of a controller that responds to actual traffic demand in real time, with the objective of minimizing vehicle delays, number of stops, etc. Fuzzy logic provides the potential for development of a system that would address these needs; The objective of this research is to design and evaluate a fuzzy logic based controller for traffic intersections that is adaptive to traffic demand. The design uses the standard input traffic flow parameters generated by existing loop detectors. The outcome of this research is a traffic controller that is very responsive to real-time traffic flow for various traffic simulations, including both recurring and non-recurring conditions. Evaluation of the performance of the system is based on minimization of delay and the number of stops. The performance of the fuzzy controller is compared to that of a pretimed controller with the help of traffic packages NETSIM & SOAP-84.