In this dissertation, a cyber-physical system called MIDAS (Managing Interacting Demand And Supply) has been developed, where the “supply” refers to the transportation infrastructure including traffic controls while the “demand” refers to its dynamic traffic loads. The strength of MIDAS lies in its ability to proactively control and manage mixed vehicular traffic, having various levels of autonomy, through traffic intersections. Using real-time traffic control algorithms MIDAS minimizes wait times, congestion, and travel times on existing roadways. For traffic engineers, efficient control of complicated traffic movements used at diamond interchanges (DI), which interface streets with freeways, is challenging for normal human driven vehicular traffic, let alone for communicationally-connected vehicles (CVs) due to stochastic demand and uncertainties. This dissertation first develops a proactive traffic control algorithm, MIDAS, using forward-recursion dynamic programming (DP), for scheduling large set of traffic movements of non-connected vehicles and CVs at the DIs, over a finite-time horizon. MIDAS captures measurements from fixed detectors and captures Lagrangian measurements from CVs, to estimate link travel times, arrival times and turning movements. Simulation study shows MIDAS’ outperforms (a) a current optimal state-of-art optimal fixed-cycle time control scheme, and (b) a state-of-art traffic adaptive cycle-free scheme. Subsequently, this dissertation addresses the challenges of improving the road capacity by platooning fully autonomous vehicles (AVs), resulting in smaller headways and greater road utilization. With the MIDAS AI (Autonomous Intersection) control, an effective platooning strategy is developed, and optimal release sequence of AVs is determined using a new forward-recursive DP that minimizes the time-loss delays of AVs. MIDAS AI evaluates the DP decisions every second and communicates optimal actions to the AVs. Although MIDAS AI’s exact DP achieves optimal solution in almost real-time compared to other exact algorithms, it suffers from scalability. To address this challenge, the dissertation then develops MIDAS RAIC (Reinforced Autonomous Intersection Control), a deep reinforcement learning based real-time dynamic traffic control system for AVs at an intersection. Simulation results show the proposed deep Q-learning architecture trains MIDAS RAIC to learn a near-optimal policy that minimizes the total cumulative time loss delay and performs nearly as well as the MIDAS AI.

The emerging multimodal mobility as a service (MaaS) and connected and automated mobility (CAM) are expected to improve individual travel experience and entire transportation system performance in various aspects, such as convenience, safety, and reliability. There have been extensive efforts in the literature devoted to enhancing existing and developing new methodologies and tools to investigate the impacts and potentials of CAM systems. Due to the hierarchical nature of CAM systems and associated intrinsic correlated human factors and physical infrastructures from various resolutions, simply considering components across different levels into a single model may be practically infeasible and computationally prohibitive in operation and decision stages. One of the greatest challenges in existing studies is to construct a theoretically sound and computationally efficient architecture such that CAM system modeling can be performed in an inherently consistent cross-resolution manner. This research aims to contribute to the modeling of CAM systems on layered transportation networks, with a special focus on the following three aspects: (1) layered CAM system architecture with a tight network and modeling consistency, in which different levels of tasks can be efficiently performed at dedicated layers; (2) cross-resolution traffic state estimation in CAM systems using heterogeneous observations; and (3) integrated city logistics operation optimization in CAM for improving system performance.

Travel time is the main transportation system performance measure used by the planning community to evaluate the impacts of traffic congestion on infrastructure investment projects and policy development plans. Planners rely on the travel demand model tool estimates for the selection and prioritization of critical and sensitive projects to meet the fiscally constraint requirements imposed by the Federal Highway Administration (FHWA) on their transportation improvement programs (TIP). While travel demand model estimates have been successfully implemented in the evaluation of project scenarios or alternatives, the application of the methods used in the travel demand model to generate these estimates continues to present a critical challenge, particularly to modelers who have to produce a validated model upon which traffic predictions can be made. The various volume-delay functions (VDFs) including the Bureau of Public Roads (BPR) function, used in the travel demand model to relate traffic volume to travel time, are developed based on system-wide attributes. BPR function in its polynomial form is computationally efficient and simple for implementation in a transport planning software. The planning community has long recognized that the BPR function cannot capture traffic flow dynamics and queue evolution processes. Besides, it has difficulties in using the average travel time measure to describe an oversaturated bottleneck with high density but low throughput. This dissertation aims to propose a simplified and yet effective point-queue based modeling approach built on the cumulative vehicle arrival concept, and the polynomial equation formula, based on Newell’s method, to estimate travel time at a corridor level using real-world speed and count measurements. A traffic state estimation (TSE) method is also proposed to characterize data into various states, such as congested state and uncongested state, using Markov Chain to capture current traffic pattern and Bayesian Classifier to infer congestion effects. As the testbed for the case study, the research selects the Phoenix freeway corridor with year-round traffic data collected from embedded traffic loop detectors. The results and effectiveness of the proposed methods are discussed to shed light on the calibration of link performance function, which is an analytical building block for system-wide performance evaluation.

Short-notice disasters such as hurricanes involve uncertainties in many facets, from the time of its occurrence to its impacts’ magnitude. Failure to incorporate these uncertainties can affect the effectiveness of the emergency responses. In the case of a hurricane event, uncertainties and corresponding impacts during a storm event can quickly cascade. Over the past decades, various storm forecast models have been developed to predict the storm uncertainties; however, access to the usage of these models is limited. Hence, as the first part of this research, a data-driven simulation model is developed with aim to generate spatial-temporal storm predicted hazards for each possible hurricane track modeled. The simulation model identifies a means to represent uncertainty in storm’s movement and its associated potential hazards in the form of probabilistic scenarios tree where each branch is associated with scenario-level storm track and weather profile. Storm hazards, such as strong winds, torrential rain, and storm surges, can inflict significant damage on the road network and affect the population’s ability to move during the storm event. A cascading network failure algorithm is introduced in the second part of the research. The algorithm takes the scenario-level storm hazards to predict uncertainties in mobility states over the storm event. In the third part of the research, a methodology is proposed to generate a sequence of actions that simultaneously solve the evacuation flow scheduling and suggested routes which minimize the total flow time, or the makespan, for the evacuation process from origins to destinations in the resulting stochastic time-dependent network. The methodology is implemented for the 2017 Hurricane Irma case study to recommend an evacuation policy for Manatee County, FL. The results are compared with evacuation plans for assumed scenarios; the research suggests that evacuation recommendations that are based on single scenarios reduce the effectiveness of the evacuation procedure. The overall contributions of the research presented here are new methodologies to: (1) predict and visualize the spatial-temporal impacts of an oncoming storm event, (2) predict uncertainties in the impacts to transportation infrastructure and mobility, and (3) determine the quickest evacuation schedule and routes under the uncertainties within the resulting stochastic transportation networks.

Gathering the necessary information required to tackle traffic congestion problems is generally time consuming and challenging but is an important part of city planners’ work. The purpose of this paper is to describe the methodology used when analyzing potential solutions for the Arizona State Route 89A and Highway 179 roundabout in Sedona, Arizona; which is currently experiencing significant congestion. The oversaturated condition is typically applied to signalized intersections but its application to roundabouts requires further exploration for future management of similar transportation systems. The accompanying Quick Estimation and Simulation model (QESM) spreadsheet was calibrated using an iterative process to optimize its level of adaptability to various scenarios. This microsimulation modeling program can be used to predict the outcome of possible roadway improvements aimed at decreasing traffic congestion. The information provided in this paper helps users understand traffic system problems, as a primary to visual simulation programs.

This dissertation investigates congestion mitigation during the ingress of a planned special event (PSE). PSEs would impact the regular operation of the transportation system within certain time periods due to increased travel demand or reduced capacities on certain road segments. For individual attendees, cruising for parking during a PSE could be a struggle given the severe congestion and scarcity of parking spaces in the network. With the development of smartphones-based ridesharing services such as Uber/Lyft, more and more attendees are turning to ridesharing rather than driving by themselves. This study explores congestion mitigation during a planned special event considering parking, ridesharing and network configuration from both attendees and planner’s perspectives.

Parking availability (occupancy of parking facility) information is the fundamental building block for both travelers and planners to make parking-related decisions. It is highly valued by travelers and is one of the most important inputs to many parking models. This dissertation proposes a model-based practical framework to predict future occupancy from historical occupancy data alone. The framework consists of two modules: estimation of model parameters, and occupancy prediction. At the core of the predictive framework, a queuing model is employed to describe the stochastic occupancy change of a parking facility.

From an attendee’s perspective, the probability of finding parking at a particular parking facility is more treasured than occupancy information for parking search. However, it is hard to estimate parking probabilities even with accurate occupancy data in a dynamic environment. In the second part of this dissertation, taking one step further, the idea of introducing learning algorithms into parking guidance and information systems that employ a central server is investigated, in order to provide estimated optimal parking searching strategies to travelers. With the help of the Markov Decision Process (MDP), the parking searching process on a network with uncertain parking availabilities can be modeled and analyzed.

Finally, from a planner’s perspective, a bi-level model is proposed to generate a comprehensive PSE traffic management plan considering parking, ridesharing and route recommendations at the same time. The upper level is an optimization model aiming to minimize total travel time experienced by travelers. In the lower level, a link transmission model incorporating parking and ridesharing is used to evaluate decisions from and provide feedback to the upper level. A congestion relief algorithm is proposed and tested on a real-world network.