Calibration and Verification of SHAKER, a Deterministic Toll Plaza Simulation Model

Abstract

A great deal of research has been conducted on Central Florida toll roads to better understand the characteristics of the tolling operation. In this paper, the use of the toll plaza queuing model SHAKER is calibrated to replicate real life toll facility traffic conditions. SHAKER is a deterministic queuing model based on classical physics equations that determines a plaza’s maximum hourly throughput by assigning vehicles to lanes based on queuing conditions. An extensive field study on the Florida’s Turnpike Enterprise network is conducted to determine the parameters affecting toll lane capacities. Data including traffic characteristics, vehicle characteristics, and toll plaza characteristics was collected and extracted at different toll locations, configurations, and pricing. From this data, periods of constant queuing were pinpointed and from those periods, factors such as demand, throughput, service time, etc., were extracted. Using the extracted parameters, SHAKER was then calibrated and validated to estimate the capacity of four different toll plazas along the Florida Turnpike Network.