Effect of Emergency Interrupt Complexity on the Performance of Adaptive Network Based Fuzzy Inference Traffic Light Control System

Abstract

ANFIS controller is an advance technique of controlling in Traffic Light Control System (TLCS) which adjusts signal timing parameters in response to real time traffic flow fluctuations. However, the performance of ANFIS controller has not been investigated in an emergency environment. Hence, this paper investigates the effect of emergency lane sensor signal interrupt complexity on the performance of an Adaptive Network Based Fuzzy Inference System (ANFIS) TLCS. The cross roads junction with two lanes per road was considered. One Pedestrian and one Railway lane were considered as emergency lanes. One Traffic Light (TL) was used to control vehicle on each road. ANFIS-TLCS was simulated using graphic user interface tool of the MATLAB. The GUI was simulated for the four different cases of emergency interrupt complexity at some specific simulation periods and the preset number of vehicles for each lane using slide button. Performance of the ANFIS controller was tested for: no, more and most complexity emergency Interrupt cases using Cost Efficiency (CE) as a performance metric. The results obtained showed that ANFIS controller performed differently in all tested cases and worse as the complexity increases but performed relatively equal and better at a higher simulation period regardless of the interrupt complexity. Hence, ANFIS controller is recommended as a better Traffic Light controlling technique regardless of any complexity at the road junction.

Keywords: Cross road, Traffic Light, Emergency Interrupt Complexity, ANFIS Object, and Cost Efficiency