Doctor of Philosophy (PhD)
First Advisor's Name
Dr. Mohammed Hadi
First Advisor's Committee Title
Second Advisor's Name
Dr. Albert Gan
Second Advisor's Committee Title
Third Advisor's Name
Dr. Xia Jin
Third Advisor's Committee Title
Fourth Advisor's Name
Dr. Priyanka Alluri
Fourth Advisor's Committee Title
Fifth Advisor's Name
Dr. Wensong Wu
Fifth Advisor's Committee Title
Sixth Advisor's Name
Dr. Yan Xiao
Sixth Advisor's Committee Title
Coordinated Traffic Operation, Traffic Management, Incidents, Signal Control, Machine Learning, Optimization, Modeling
Date of Defense
Non-recurrent congestion is one of the critical sources of congestion on the highway. In particular, traffic incidents create congestion in unexpected times and places that travelers do not prepare for. During incidents on freeways, route diversion has been proven to be a useful tactic to mitigate non-recurrent congestion. However, the capacity constraints created by the signals on the alternative routes put limits on the diversion process since the typical time-of-day signal control cannot handle the sudden increase in the traffic on the arterials due to diversion. Thus, there is a need for proactive strategies for the management of the diversion routes performance and for coordinated freeway and arterial (CFA) operation during incidents on the freeway. Proactive strategies provide better opportunities for both the agency and the traveler to make and implement decisions to improve performance.
This dissertation develops a methodology for the performance management of diversion routes through integrating freeway and arterials operation during incidents on the freeway. The methodology includes the identification of potential diversion routes for freeway incidents and the generation and implementation of special signal plans under different incident and traffic conditions. The study utilizes machine learning, data analytics, multi-resolution modeling, and multi-objective optimization for this purpose. A data analytic approach based on the long short term memory (LSTM) deep neural network method is used to predict the utilized alternative routes dynamically using incident attributes and traffic status on the freeway and travel time on both the freeway and alternative routes during the incident. Then, a combination of clustering analysis, multi- resolution modeling (MRM), and multi-objective optimization techniques are used to develop and activate special signal plans on the identified alternative routes. The developed methods use data from different sources, including connected vehicle (CV) data and high- resolution controller (HRC) data for congestion patterns identification at the critical intersections on the alternative routes and signal plans generation. The results indicate that implementing signal timing plans to better accommodate the diverted traffic can improve the performance of the diverted traffic without significantly deteriorating other movements' performance at the intersection. The findings show the importance of using data from emerging sources in developing plans to improve the performance of the diversion routes and ensure CFA operation with higher effectiveness.
Saha, Rajib Chandra, "Integrated Approach for Diversion Route Performance Management during Incidents" (2021). FIU Electronic Theses and Dissertations. 4626.
In Copyright. URI: http://rightsstatements.org/vocab/InC/1.0/
This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s).