Estimating Demand Volume for Signalized Corridors Under Oversaturated Conditions Using Aggregated Probe Vehicle Speed Data

Abstract

Measuring demand directly with vehicle sensors is not possible when demand is larger than capacity for an extended period, as the queue grows beyond the sensor, and the flow measurements at a given point cannot exceed the capacity of the section. The main objective of this study is to develop methods that could be implemented in practice based on readily available data. To this end, two methods are proposed: an innovative method based on shockwave theory and the volume delay function adapted from the Highway Capacity Manual (7th edition). Both methods primarily rely on probe vehicle speeds (e.g., from INRIX) as the input data and the capacity of the segment or bottleneck being analyzed. Probe vehicle data are used to determine the critical times when the queue reaches the end or beginning of a road segment. From these critical times, the shockwave speed for the boundary between congested (high density) traffic and arriving (low density) traffic is estimated. The proposed methods are tested with simulation data generated in VISSIM and validated based on volume data from the field. The field data are collected from a congested arterial in Virginia Beach, VA, and include the ground truth volumes and INRIX speed data aggregated at one-minute intervals. The results show that both methods are effective for estimating the demand volume and produce less than 4% error when tested with field data.