Optimal Spacing of Traffic Counting Stations on a High-Volume Highway Based on Queuing Models

Abstract

This study aims to investigate the optimal spacing of the counting stations in order to get more detailed and useful traffic information (e.g. speed, density, occupancy, and flow rates.) on high-volume highways. Assuming the detector as a server in a one-queue system, and the maximum number of vehicles between two consecutive detectors as the buffer, the layout problem then converts to find out the finite buffer capacity of the queue system. Therefore, the M/M/1/K queuing model with Poisson arrival processes and exponentially distributed service time was chosen to solve the above problems. The analysis shows that two factors, such as the blocking probability and the traffic intensity, affect the value of the buffer size. The queue model can best describe the optimal spacing of the counting stations under high traffic volume conditions, especially when the traffic intensity varies from 0.6 to 0.85. In addition, after determining the layout spacing of detectors, the problems regarding using traffic volume data collected by these detectors to estimate the traffic states was equally considered. The results indicate that our methods can estimate traffic states more intuitively and conveniently.