THE DEVELOPMENT OF FUZZY CONTROL ALGORITHMS FOR A SIGNALISED PEDESTRIAN CROSSING : FUZZY SIMILARITY AND CALIBRATION OF MEMBERSHIP FUNCTIONS

Fuzzy logic has been tested in controlling traffic signals since the 1970s and has been shown to decrease delays when compared to traditional vehicle-actuated control. The ability of fuzzy logic to handle complex, dynamic and inexact situations is one of the reasons it has been applied to traffic control. Using simulation, fuzzy logic has been tested at simple pedestrian crossings with push-buttons and vehicle-actuated signal control. The problems with pedestrian crossings are mainly increased vehicle delay, when pedestrians activate the push-button but do not wait for the signal to turn green and decreased traffic safety. In the optimization of traffic signals it is important to keep pedestrian waiting time as short as possible while at the same time minimizing vehicle delay and the danger of rear-end collisions. Simulation results have shown that fuzzy logic is one of the very promising new methods of optimizing the functioning of traffic signals. This study was a part of an international research project FUSICO (FUzzy SIgnal COntrol) started in 1996. The aim of FUSICO is to study the opportunities of fuzzy logic in controlling traffic > >