Capacity Analysis of Traffic-Actuated Intersections

This thesis proposes two models that estimate the capacity of an intersection with actuated control. The capacity of an approach to or a lane group of the intersection is a function of the saturation flow rate, the green time allocated to this approach or lane group, and the cycle length of the intersection. The Minimum Delay Model estimates the green times and the cycle lengths from flow rates, minimizing the total delay at the intersection. Parameters, the ratio of green extension period to queue service time specific to each approach or lane group, are introduced into this model. The parameters depend on the distribution of arrivals of vehicles at the intersection. The Hybrid Model combines the deterministic queuing model that estimates the queue service time and a theoretical model that estimates the green extension period from the unit extension, the flow rate, the speed limit of the approach, and the detector length. A method converting the left-turn traffic volume to equivalent through volume is developed. The method is applied to estimating the capacity of intersections with permitted left-turn phases. The Minimum Delay Model and the Hybrid Model are validated at the intersection level by comparing the estimations of effective green ratios with those simulated by MITSIMLab. These two models are also validated at the network level with real data from Irvine, California. The results show that both the Minimum Delay Model and the Hybrid model are appropriate for estimating capacity of intersections with actuated control. The Minimum Delay Model is also suitable for estimating capacity of intersections with adaptive control. The Emulation Model is applicable to off-line mesoscopic dynamic traffic assignment. Thesis Supervisor: Moshe E. Ben-Akiva Edmund K. Turner Professor of Civil and Environmental Engineering Thesis Supervisor: Haris N. Koutsopoulos Volpe National Transportation Systems Center