Control of a complex traffic junction using fuzzy inference

Abstract

Fuzzy controllers have been widely used in many consumer products and industrial applications with success over the past two decades. For traffic control, however, fuzzy controllers have not been widely applied. This paper proposes a fuzzy traffic lights controller to be used at a complex traffic junction in the middle of Kuala Lumpur city, Malaysia. The proposed fuzzy traffic lights controller is capable of communicating with neighbor junctions and manages phase sequences and phase lengths adaptively. A real case study of a complex traffic junction is simulated having 4 intersections. Average flow density, average delay time and link overflow of all the 4 intersections are used as performance indices when comparing the fuzzy controller with two other existing traffic lights controllers in Malaysia, namely the preset-cycle time and vehicle-actuated controllers. A simulator has been developed to show the effectiveness of the fuzzy traffic controller which can also be used for teaching purposes.