Research

Agent-Based Approach for Integrated Driver and Traveler
     Behavior Modeling

A project team under the guidance of Assistant Professor of Civil and Environmental Engineering Lei Zhang (Principal Investigator) has been awarded a grant from the FHWA, to apply agent-based modeling and simulation to analyses of driver behavior and transportation systems management, and in providing insights for FHWA capital investments. The Department of Civil and Environmental Engineering (CEE) at the University of Maryland (UMD) and its research team partners, including UMD’s Traffic Operations and Safety Lab (TOSL), National Center for Smart Growth (NCSG), Center for Complexity in Business (CCB) and the Transportation Center and Human Performance Laboratory at the University of Massachusetts (UMass) will draw on a wealth of experience to meet the challenges of agent-based modeling and simulation (ABMS) in transportation.

The goal of this project is to use agent-based modeling to understand transportations systems, which consist of numerous intelligent agents such as travelers, drivers, and vehicles that interact with one another on various time scales in urban and regional systems, producing important and often complex system level patterns, such as travel demand and congestion. This project is a direct result of a recent Federal Highway Administration (FHWA) Exploratory Advanced Research (EAR) workshop titled “Agent-Based Modeling and Simulation” (May 4, 2010; McLean, Virginia), which CCB Research Director William Rand presented at. This workshop identified existing and recommended future applications of agent-based methods in transportation.

This project will make use of agent-based modeling and simulation in transportation to develop innovative methods in: (1) Improving our understanding of driver and traveler behavior; (2) Enhancing transportation systems management; and (3) Providing new insights for capital investments. Moreover, this project will address several technical challenges including: (1) Data requirements and needs; (2) Agent behavior specification, estimation, and validation; and (3) Software platform for model implementation and applications.

Building on this vision, the research team will develop an innovative agent-based approach for integrated driver and traveler behavior modeling with applications for transportation systems management, capital investment evaluation, transportation planning, and beyond. The proposed scope of work focuses on agent decision types including en-route diversions, pre-trip route choice, departure time choice, and mode choice, which provide the crucial linkages between traditional traffic simulation, travel demand and innovative agent-based models.