Civil and Systems Engineering Seminar: Brian Park

Description

Brian Park, professor of civil engineering and systems engineering and a member of the Link Lab at the University of Virginia, will give a talk titled "Transformative Traffic Management Harnessed by Proactive Route Guidance" as a Civil and Systems Engineering Seminar.

Abstract:

The route guidance system is one of the most effective ways of reducing traffic congestion. This talk presents transformative traffic management using a proactive user optimum-oriented route guidance system that incorporates individual users' preferences to achieve users' better satisfaction and transportation system performance improvement.

To ensure the route guidance system accurately captures each driver's route choice preference, the traditional discrete choice model, mixed logit model, support vector machine, and multi-task linear model adaptation (MT-LinAdapt) were implemented using the stated preference dataset collected from 102 participants. The evaluation results showed that MT-LinAdapt outperformed the other methods. In addition, a deep neural network approach, namely, hybrid-transformer-LSTM, was designed to model route choice behaviors using detailed network features and real-world route choices.

The proposed transformative traffic management using a proactive user optimum-oriented route guidance system was evaluated using a commonly used Sioux Falls network and user population whose preferences were synthesized from surveyed participants. The proposed approach demonstrated advantageous performance in terms of users' satisfaction (up to 22% more satisfied users), system mobility and sustainability (up to 10% travel time reduction and up to 42% delay reduction), and future traffic conditions estimation (up to 70% links having more accurate volume estimation), compared to a user equilibrium solution.