An intelligent highway tollgate queue selector for improving server utilization and vehicle waiting time

Abstract

On a highway setup, a vehicle will most probably use a tollgate server that has the shortest queue thinking that it is the fastest exit. In this paper, an intelligent highway tollgate queue selector using fuzzy logic is proposed and simulated in Matlab SimEvents. Its aim is to automatically select the most appropriate tollgate server for a vehicle to ensure the shortest waiting time while trying to balance the server's utilization. Two policies are considered in this study, namely: (1) Shortest Queue (SQ) and (2) Fuzzy Logic-Controlled Queue (FLCQ). Results indicate that the FLCQ policies reduce the average waiting time and queue length by approximately 50% of those obtained from the SQ policy while guaranteeing an equal utilization among available servers. These findings are valid for light and heavy, homogeneous and nonhomogeneous vehicle arrivals. To further improve the decision making of the fuzzy logic controller, traffic flow information collected by remote road-side units can be exploited to allow the control of various system parameters (e.g., service time) in advance.