Integration of Predictive Routing Information with Dynamic Traffic Signal Control
Abstract
This thesis explores the integration of predictive routing information available under the Intelligent Vehicle Highway System (IVHS) with dynamic traffic signal control. This exploration was motivated by recent advances in both signal processing and computational technology. The first portion of the thesis develops the theoretical basis for the Predictive Routing Information Signal Timing INtEgration (PRISTINE) model. PRISTINE explicitly uses the predictive routing information available under IVHS as well as considering the effects of queuing and congestion and compensating for them in setting the traffic signal control plan. The thesis develops a new methodology for using spanning trees to determine the offsets in the network and draws several theoretical results from this premise. The thesis also develops a Queue Effects Model (QEM) that explicitly considers the effects of bulk arrivals on average delay and probability of stopping at an intersection; the Queue Effects Model requires only the average arrival rate and the first two moments of platoon size, all of which can be collected using existing technology. The thesis uses the Queue Effects and spanning tree models as the basis for PRISTINE. Two methods are developed for selecting the splits and cycle time in PRISTINE.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 01, 1994
- Accession Number
- ADA280077
Entities
People
- Richard C. Staats
Organizations
- Massachusetts Institute of Technology