| Description |
The Active Traffic and Demand Management (ATDM) initiative aims to integrate various management strategies and control measures so as to achieve the mobility, environment and sustainability goals. To support the active monitoring and management of real-world complex traffic conditions, the first objective of this dissertation is to develop a travel time reliability estimation and prediction methodology that can provide informed decisions for the management and operation agencies and travelers. A systematic modeling framework was developed to consider a corridor with multiple bottlenecks, and a series of close-form formulas was derived to quantify the travel time distribution under both stochastic demand and capacity, with possible on-ramp and off-ramp flow changes. Traffic state estimation techniques are often used to guide operational management decisions, and accurate traffic estimates are critically needed in ATDM applications designed for reducing instability, volatility and emissions in the transportation system. By capturing the essential forward and backward wave propagation characteristics under possible random measurement errors, this dissertation proposes a unified representation with a simple but theoretically sound explanation for traffic observations under free-flow, congested and dynamic transient conditions. This study also presents a linear programming model to quantify the value of traffic measurements, in a heterogeneous data environment with fixed sensors, Bluetooth readers and GPS sensors. It is important to design comprehensive traffic control measures that can systematically address deteriorating congestion and environmental issues. To better evaluate and assess the mobility and environmental benefits of the transportation improvement plans, this dissertation also discusses a cross-resolution modeling framework for integrating a microscopic emission model with the existing mesoscopic traffic simulation model. A simplified car-following model-based vehicle trajectory construction method is used to generate the high-resolution vehicle trajectory profiles and resulting emission output. In addition, this dissertation discusses a number of important issues for a cloud computing-based software system implementation. A prototype of a reliability-based traveler information provision and dissemination system is developed to offer a rich set of travel reliability information for the general public and traffic management and planning organizations. |
