The 3-Minute Thesis (3MT) Competition invites participants to present their research within three minutes in a way that the workshop attendees from wide research areas in Transportation can absorb the objectives and key components of the presented work. The competition is open to graduate students attending the TRB annual meeting. The OPG YMC will follow guidelines and judging criteria established by the University of Queensland, the founder of the 3MT competition. The workshop aims to provide an excellent opportunity for young members to consolidate their ideas, so they can present them concisely to an audience specialized in various areas in Transportation. It will help students gain visibility of their research as well.
Eligibility and Submission Guideline
The competition is open to graduate students conducting research in all areas of Transportation who have made significant progress towards their degrees. Students need to apply to the 3MT workshop by submitting (1) an abstract on their research (up to 350 words) and (2) their CVs as a single PDF. The submission window will be open until November 1, 2018. Applications will be reviewed by the judges before the annual meeting and selected participants will be notified to prepare their presentations. Selected speakers will have three minutes to present a compelling oration on their dissertation and its significance. The presentation guidelines will be announced through email notifications.
The presentations will be judged by a panel of experts and an anonymous poll of the audience. The experts are selected from TRB standing committee chairs and members. The judgement will be based on the qualifications including the clarity of the presentation and effectiveness of the communication skills along with scientific and technical quality.
The workshop will have three winners. Top two presentations (winner and runner-up) will be identified by the judges and an Audience Choice will be recognized as well.
An important aspect of the competition is the Audience Choice award. Following all presentations, the audience is asked to vote on who they thought gave the most convincing 3MT presentation based on the outlined criteria. Details of the competitors will be provided to the audience along with voting papers to select the number which corresponds to their favorite presenter. A finalist can be the winner or runner-up and still receive the Audience Choice award.
The three winners will be recognized at the end of the workshop and receive an award certificate from TRB.
Timothy B. Baughman, P.E., Institute for Transportation Research and Education
Robert L. Bertini, Ph.D., P.E., F.ASCE, F.ITE, SMIEEE, University of South Florida
Frank Broen, Teach America Corporation
Marc Butorac, PE, PTOE, PMP, Kittelson & Associates
Xiaoyue Cathy Liu, Ph.D., P.E., University of Utah
Utilization Management and Pricing of Urban Facilities Under Uncertain Demand and User Decisions
Amir Mirheli, State University of New York (SUNY)Show Abstract
Excessive demands for urban facilities (e.g., parking lots, charging stations) and shortage in supplies contributes to indirect socio-economic impacts. Inefficient strategies to manage facilities can yield minor positive returns, but may not lead to long-term steady solutions. While planning to mitigate congestion, it is beneficial to reflect diversity of users’ behaviors and their travel choices along with addressing facility manager’s objectives. This dissertation first develops a stochastic dynamic parking management model, under competitive user-agency perceptions and uncertain user demand and parking occupancy, to simultaneously minimize the total travelers’ costs and maximize parking agency’s revenue. The problem is formulated into a dynamic programming model and solved using a stochastic look-ahead technique based on Monte Carlo tree search algorithm to determine optimal actions on parking price assignment and spot allocation over time. Then, in the second study, a management framework is developed to incorporates a dynamic pricing scheme into facility location design for electric vehicle charging infrastructure. The objective is to determine the optimal location and capacity of charging facilities in the transportation network and the optimal pricing based on demand. The problem is formulated as a bi-level optimization program that minimizes the total costs including infrastructure investments and operation costs of facilities under user-equilibrium flows. Numerical experiments in both studies confirm that the proposed methodologies solve the problem efficiently.
Assessment of Tree Canopy Effects Overtop Low-Volume Roadways
Andrea Horn, Ohio UniversityShow Abstract
Located in the Great Lakes region of the US, the State
of Ohio is captivating with its
Improving the Quality Management of Pavement Condition Data
Bahareh Bazargani, Iowa State UniversityShow Abstract
Improving Quality Management of Pavement Condition
Life-Cycle Thinking in Pavement Design Decision Making
Chaitanya Ganesh Bhat, Michigan Technological UniversityShow Abstract
As the infrastructure industry moves towards communicating the environmental impacts of products and processes through Life-Cycle Assessment (LCA) outcomes, there is a growing need to develop methods that support the informed use of mid-point impact indicators such as Global Warming Potential (GWP), in the decision-making process. This research presents a framework that will aid decision-makers inform the design of pavement systems by explicitly accounting for mid-point environmental impact indicators along with performance requirements and costs. An illustration of the framework is provided for the design of a benchmark asphalt mixture, considering its GWP, production costs, Reclaimed Asphalt Pavement (RAP) content in design, and its performance characteristics in an asphalt pavement with respect to thermal cracking, rutting and alligator cracking. Trade-offs between the competing multiple objectives of pavement performance, costs, and GWP are characterized using the method of Lagrange multipliers. Next, a life-cycle thinking approach is proposed to holistically identify ways to reduce the GWP of the mixture without reducing performance. The inverse trade-off relationship between thermal cracking and GWP is improved by associating it with the transportation distance of RAP from source to plant. The outcomes of the analysis provide decision-makers with guidelines on margins of tolerance and indicate the possibility of identifying ways to reduce the environmental impacts of a design without impacting performance negatively.
Development, Assessment, and Modeling of Self-Healing Systems for Cementitious Materials
Jialuo He, Washington State UniversityShow Abstract
Cracking is known to be the most challenging problem for the life-cycle performance of cementitious materials, which are the most widely used construction materials in pavements and highways. The annual costof repair and maintenance is almost the same as that of new construction. My research is mainly focused on the development self-healing systems for cementitious materials to reduce the cost of repair and extend the service life of concrete structures. One is urea-formaldehyde microcapsule- based self-healing system and the other self-healing system is encapsulated lightweight aggregates impregnated with healing agents. Microcapsule-based self-healing system was designated to heal the cracks created during the curing process due to early-age shrinkage, and both the microcapsules and encapsulated lightweight aggregates were designated to heal the cracks created throughout the service life. Early-age shrinkage, compressive strength, gas permeability, surface resistivity, rapid chloride migration test and mercury intrusion porosimetry test were conducted to evaluate the healing performance of the microcapsule-based self-healing system. Besides, material characterizations such XRD, SEM-EDX and TGA were conducted to investigate the healing products. Four parameters of the encapsulating materials (i.e., curing temperature, curing time, the concentration of PVA and the concentration of ethylene glycol/water-based epoxy) were optimized based on the response surface method, in which water contact angle and adhesive strength were tested. Water absorption, gas permeability, compressive strength and uniaxial tensile strength were tested to evaluate the healing efficiency of the self-healing system of the encapsulated lightweight aggregates impregnated with healing agents. In addition, a micromechanical model was developed to simulate the mechanical regain. Material characterizations such XRD, SEM-EDX and TGA were also conducted. The resistance of freeze-thaw actions of self- consolidating concrete embedded with both self-healing systems were investigated. Computed Tomography was employed to reconstruct the 3-D image of the microstructure and the probabilistic damage model was developed to predict the service life of such self-healing concrete. The addition of microcapsule-based self-healing system mitigated 25% of the early-age shrinkage, and the encapsulated lightweight aggregates showed over 50% of the mechanical regain. The addition of both self-healing system improved 25% of the resistance of freeze-thaw actions.
Developing a Sustainable Pavement Management Plan: Tradeoffs in Road Condition, User Costs, and Greenhouse Gas Emissions
Jojo France-Mensah, University of Texas, AustinShow Abstract
Highway Maintenance and Rehabilitation (M&R)
planning is a multi-faceted decision-making process which requires the
consideration of multiple but often conflicting objectives. Traditionally, State
Highway Agencies (SHAs) have focused on maximizing the condition of a pavement
network or minimizing agency costs needed to keep the network at an acceptable
physical and functional level of service. However, other objectives like
minimizing the road user costs and greenhouse gas (GHG) emissions are seldom
factored into the M&R decision-making process. In this paper, a mixed
integer program that accounts for these three objectives is presented. The
proposed model is then implemented on a subset network in Texas consisting of
100 pavement sections. The results demonstrate the Pareto-optimal relationships
that exist between the road condition improvements, road user costs, and GHG
emissions. Notably, there is a positive relationship between condition score
improvements and the observed marginal GHG emissions from the case network
implementation. Furthermore, this network’s results indicate that GHG emissions
from materials and construction contribute to the most significant proportion of
the total estimated GHG emissions. The results of the sensitivity analysis
confirm the observed trends in the network case examined and provide valuable
insights concerning the impact of variability in the budget size. The principal
contribution of this study is in providing an approach for highway
decision-makers to perform trade-offs in road condition improvements, road user
costs, and GHG emissions of alternate pavement M&R programs or
Corrosion of Corrugated Metal Pipe in Kansas: A Longitudinal Study
Luke Augustine, Kansas State UniversityShow Abstract
Corrugated Metal Pipe (CMP) in the transportation
industry is most commonly used for stormwater management. CMP is very
susceptible to corrosion, similar to all buried metallic elements. Studies have
been conducted by several departments of transportation to evaluate the
corrosion of CMP; few have conducted follow-up studies. In this study, 80 CMP
were examined in 2018, 41 of which were previously examined in 1990. The
objective of this study was to evaluate if CMP policy changes implemented by the
Kansas Department of Transportation in 1975 and 2001, as well as the
recommendations made as a result of a study conducted in 1990, impacted CMP
corrosion performance. The results from 2018 were compared to the same CMP
results from the 1990 study to support the research findings. It was determined
that the invert is the most significant when assessing corrosion in CMP; the
invert, defined as the bottom of the interior of the CMP, showed the highest
variability in corrosivity when compared with age. The results also indicated
that aluminized CMP may be more resistant to corrosion than galvanized CMP. The
findings of this study will be used to help shape the long-term CMP policy at
the Kansas Department of Transportation.
A Laboratory-Based Procedure for Testing and Evaluating Pavement Markings
Maged Mohamed, University of IdahoShow Abstract
Drivers rely on pavement markings to maintain a safe
road path especially during nighttime and challenging weather conditions.
Despite the growing demand for pavement marking materials that facilitate road
driving by providing increased visibility, there is concern about the durability
and long-term weatherability of these products. Current performance evaluation
methodology using National Transportation Product Evaluation Program (NTPEP)
test-deck protocols consume significant time and effort. The objective of this
study was to model the deterioration of pavement markings in the field and to
develop an accelerated laboratory-based procedure that could evaluate and study
the performance of different pavement marking products and shorten the
evaluation duration. For this study, field data for pavement markings from
thirty-eight sites were collected and analyzed over twelve months in the State
of Idaho across six districts with different environmental conditions. To
replicate varying traffic, snowplowing, and weather conditions in the
laboratory, a three-wheel polisher device (TWPD) and weatherometer (Q-SUN Xe-1
Xenon Arc) were employed. Comparisons between the physical measurements (i.e.,
retroreflectivity, color change, and durability) of the waterborne (nondurable)
and the thermoplastic (durable) marking materials were successfully achieved. In
the field, the results yielded a logarithmic relationship between the
retroreflectivity and the age of waterborne markings, and pavement markings in
districts subjected to higher ground snow loads deteriorated faster than those
with lower ground snow loads. The TWPD results illustrated that the
retroreflectivity, total color change, and durability of both materials
logarithmically deteriorated and changed under different loadings, except the
durability of the thermoplastic markings which followed a linear degradation
function. The artificial weathering results illustrated a linear increase in
percent retroreflectivity similar to the increase that occurs in the field after
installation along with similar long-term color change patterns. The results
indicated a significant relationship between all performance measures assessed
in the laboratory and field. This research provides guidance to federal and
state transportation agencies, particularly those sited in cold-weather regions,
who are concerned with selecting and maintaining pavement markings and to
researchers studying the importance of pavement markings as it relates to road
Computer Vision–Based Structural Health Monitoring Framework to Feature Mining of Damage for Predictive Numerical Simulations
Mehrdad Shafiei Dizaji, University of VirginiaShow Abstract
In this proposal, I have proposed two complementary
computer vision based approaches to detect defections on the surface of the
structural elements such as concrete columns, concrete beams and etc. as
Distributed Optimization and Coordinated Algorithm for Dynamic Speed Optimization of Connected and Autonomous Vehicles in Urban Street Networks
Mehrdad Tajalli, Washington State UniversityShow Abstract
Connected vehicle and other advance communication technologies create possibilities to facilitate the movement of vehicles through transportation networks and reduce their travel time. Optimizing the speed of vehicles in different network links not only yields a more efficient network capacity utilization, but also regulates the movement of vehicles to achieve a "smoother" flow of traffic. In addition, vehicles speed can be cooperatively optimized along with other traffic controllers such as traffic signals at intersections to further reduce the number of stops and prevent gridlocks. In our research, we formulated the speed optimization problem for connected and autonomous vehicles in urban street networks. Moreover, we provided a formulation to coordinate vehicles speed and traffic signal controllers in the network. The objective function of problems is maximizing the total number of completed trips in the network and harmonize vehicles speed at the same time. Since the computational complexity of the speed optimization problem in the network is very high, we developed a distributed algorithm to decompose the problem into several subproblems and solve them for smaller subnetworks. Moreover, we pushed solutions toward optimality by providing a coordination between subnetworks through communications between vehicles to vehicle and vehicles to infrastructures. As a result of distributed optimization and coordinated algorithm, both speed optimization and coordinated signal and speed optimization problems can be solved in real time for various size of transportation networks with a small optimality gap.
System Optimal DynamicTraffic Assignment
Mehrzad Mehrabipour, Washington State UniversityShow Abstract
My research studies have been dedicated to the development of decomposition and distributed approaches for system optimal dynamic traffic assignment (SODTA) problems with Cell Transmission Model (CTM) to find time-dependent traffic flows to minimize total cost in urban street transportation networks. CTM can capture shockwaves, speed, and queue properties despite traditional traffic assignment approaches. However, it creates intractable computational complexity. Exact and heuristic approaches have been developed to solve CTM-based SODTA problems. Exact algorithms are applicable to problems with limited temporal and spatial scales. Heuristic methodologies are scalable and more efficient in run-time. However, they cannot generate solutions in real-time, and they are iterative approaches with long run-times. The path-based structure is another drawback that leads to limiting the number of paths for each OD to control computational complexity and a requirements for a large amount of storage. We first developed a decomposition algorithm, featuring OD-based sub-problems, that exploits the structure of CTM-based SODTA formulations. The algorithm can find solutions for problems with significantly larger temporal and special scales with millions of decision variables as the largest CTM-based SODTA problems that is solved in the literature, based on my knowledge. The algorithm guarantees to generate feasible solutions and finding the optimal solutions in a finite number of iterations. It stores and updates cell-based flows that lead to significant computational time and memory savings compared to path-based counterparts. Moreover, the algorithm creates independent sub-problems that provide the possibility of parallelism. Then, we introduced an efficient algorithm to find near-optimal solutions to system optimal dynamic traffic assignment problems in real-time. The algorithm decomposes network-level traffic assignment problems into several intersection-level sub-problems that can be solved individually. As a result, the complexity of problems is reduced significantly, and solutions can be found in real-time. The sub-problems coordinate their decisions by exchanging information with other sub-problems and push the solutions towards global optimality. I have also worked on proposing a distributed algorithm for an integrated signal timing optimization and SODTA. The studies are also tested on real-world networks with promising results.
Model of Optimal Combination Scheduling Based on Passengers’ Willingness to Pay for the Demand of Bus Seats
Mingyang Pei, University of South FloridaShow Abstract
This paper proposes a new model of optimal combination
scheduling which has a background of the spread of high quality public
transport. Considering passengers’ willingness to have a seat on the bus, we
design a bus line with two kind of bus types, one is the ordinary bus which has
few seat and flat fare, the other is the VIP bus which have enough seats for
everyone and a higher fare. In the rule of this paper, all the passengers should
take on the bus, and they are arriving and leaving both at a uniform
distribution in the peak hours. The willingness to pay of passengers’ is
steadfast and never changes in the operation time.
Field Performance and Cost-Effectiveness of Crack Sealing in Flexible and Composite Pavements
Momen Ragab Mousa, Louisiana State UniversityShow Abstract
Crack sealing prevents the ingress of water in the pavement structure, thus preventing the weakening of the pavement and delaying its deterioration. Earlier studies indicated that sealing pavements in areas with high ground water table (GWT) prevented moisture from escaping upwards through the cracks of asphalt pavements, therefore, accelerating stripping. The objectives of this study were to (1) evaluate the field performance and cost-effectiveness of crack sealing in asphalt pavements in hot and wet climates such as Louisiana, and to develop a model that would quantify the expected benefits of crack sealing given project conditions; and (2) provide guidelines for using crack sealing to minimize moisture entrapment under crack tips, therefore, reducing stripping. To achieve the first objective, 28 control sections that were crack-sealed between 2003 and 2010 were monitored for at least 4 years. Results indicated that when compared to untreated segments, crack sealing extended Pavement Service Life (PSL) by two years. When compared with the original pavement, crack sealing extended PSL by 5.6 years, if applied at the correct time. To achieve the second objective, calibrated Finite-Element (FE) model was used to model a field experiment consisting of cracked and crack-sealed asphalt pavement sections. Sensitivity analysis was then conducted to compare between crack-sealed and unsealed sections under different GWT levels, air relative-humidity, air-temperatures, rain-intensities, and asphalt hydraulic conductivities. Results indicated that crack sealing could be applied under common rain intensities in Louisiana and any GWT depth without potential for stripping due to moisture entrapment if the hydraulic conductivity of the original pavement does not exceed 2x10-6 m/s. Yet, crack sealing should be applied after a dry period to ensure that the existing moisture in the original pavement is minimal. A non-linear regression model was developed to assist in determining whether crack sealing should be used to avoid moisture damage in a cracked pavement at a given site based on the GWT and air relative-humidity without the need for FE simulations. This can be a useful tool during the planning stage of maintenance activities.
Leveraging Autonomy in Truck Platooning to Improve Transportation Sustainability
Osman Erman Gungor, University of Illinois, Urbana ChampaignShow Abstract
Abstract Introduction of autonomous and connected trucks
(ACTs) is expected to result in drastic changes in operational characteristics
of freight shipments, which may in turn have significant impacts on efficiency,
safety, energy consumption, and infrastructure durability. One such important
change would be the formation of truck platoons which will be more feasible and
practical with the intelligent technologies existing in ACTs that enable the
connection among vehicles and between vehicles and infrastructure. Reducing
congestion, braking/accelerating and fuel efficiency are some of reported and
expected benefits of the platooning. Yet, such platooning operations may
accelerate the damage accumulation within pavement structures because their
similar lateral positions (i.e., wheel wander) within the lane. Therefore, this
study develops a platooning control strategy for a fleet of ACTs such that
trucks’ lateral shifts within a lane can be explicitly optimized to minimize
damage to the pavement, thus significantly reducing maintenance and
rehabilitation costs. The efficacy of the proposed strategy was demonstrated
through a case study. The results showed that the cost of platooning can be
reduced by approximately 0.5 15 million $/mi depending on the platoon size,
pavement thickness and traffic.
Application of a New Laboratory Compaction Method for Granular Materials Used in Transportation Infrastructure
Poura Arabali, Texas A&M UniversityShow Abstract
Compaction of granular materials minimizes air voids, increases density and shear strength, and lessens compressibility and water permeability. Impact compaction has been the most prevalent method of laboratory compaction of granular materials, also preparing specimens for different tests. Factors such as high variability of unconfined strength test in the specimens compacted with impact hammer and presence of layer interface between lifts can be downsides of this method. Moreover, impact compaction uses impact loads, while modern field compaction applies a combination of vibration, kneading, and increased pressures. This study investigates an alternative laboratory compaction method for unbound granular materials. The effects of using gyratory compactor on the engineering properties of unbound granular materials used in pavement infrastructure are studied. Gyratory compaction applies normal pressure and shear forces generated due to gyratory motions. An experimental program was performed on the specimens compacted with both Superpave gyratory compactor and impact hammer in order to compare their engineering properties. Unconfined compressive strength tests were conducted to investigate whether using gyratory compaction can enhance precision of this test. The results indicated that specimens prepared with the gyratory compactor showed higher precision in the strength test than the impact hammer. Furthermore, maximum dry density and optimum moisture content were determined by conducting impact and gyratory compaction tests. Statistical analyses indicated that the gyratory compactor resulted in slightly different maximum dry density and optimum moisture content in the studied materials. Additionally, this research investigated materials behavior under cyclic loading, used for prediction of performance of aggregate layers and pavement structure in mechanistic-empirical pavement design. Permanent deformation and resilient modulus testing and modeling, as performance-related characteristics, were performed on the specimens made with these two compaction methods. Materials showed different behavior regarding permanent strain and resilient modulus in these two series of samples. The effects of granular materials properties on the engineering behavior in the two different compaction methods were studied. An equation for compaction energy was also developed to quantify the required compaction effort using gyratory compactor, which revealed substantial differences between different materials. Gyratory compaction produces a different mechanism of compaction from the impact compaction.
Data-Driven Damage Prediction Tool for Airfield Flexible Pavements
Priyanka Sarker, University of Illinois, Urbana ChampaignShow Abstract
Unlike highway pavements which see mostly channelized traffic, airfield pavements experience higher load levels with wander (lateral movement of aircraft). Though wander reduces the number of repetitions of maximum load applied to the most heavily-trafficked pavement location, wander does not necessarily increase the pavement life and can in reality be quite damaging to the pavement system. Considering the New-Generation Aircraft (NGA) applying heavier wheel loads and aircraft wander coupled with complex gear configurations, airfield pavement engineers and designers require a better understanding of pavement rutting damage mechanisms built into improved rut prediction tools. Addressing this concern, the objective of this research was to develop a data driven rut prediction methodology to account for permanent deformation damage accumulation trends in granular base/subbase layers of airfield flexible pavements. The field data analyzed were from Construction Cycle 5 (CC5) tests conducted on instrumented airfield pavement sections, built with two different subbase materials and tested under various wheel loads applied using two different landing gear configurations, at the Federal Aviation Administration’s (FAA’s) National Airport Pavement Test Facility (NAPTF) located in Atlantic City, NJ. The majority of the surface rutting was found to be due to permanent deformations observed in the granular base and subbase layers. Critical wander locations were established and their contributions to transverse rut profiles were studied for multiple passes. By utilizing the measured multi-depth deflectometer (MDD) individual pavement layer deformations and the periodic transverse field surface profiles, a rut prediction model was developed using general linear models in the forms of power and sigmoidal function distributions to determine realistic surface profiles of the CC5 test sections. It was observed that both the power and sigmoidal models could predict accurate field surface rut profiles up to 12,000 passes. However, at higher gear/wheel passes the sigmoidal model predictions were more accurate than those of the power predicted ones. The developed rut prediction methodology is completely data driven and can take into consideration any random paths and wander patterns of an airplane traveling on a taxiway/runway to predict the shapes and magnitudes of pavement surface deformation profiles.
Traffic Metering in Urban Street Networks
Rasool Mohebifard, Washington State UniversityShow Abstract
Traffic congestion is one of the contributing factors to the excessive delay of travelers, air pollution, and fuel consumption of vehicles in an urban network. As traffic congestion increases, the network capability for processing vehicles reduces due to queue spillovers and gridlocks. In this condition, controlling the number of entry vehicles to the congested areas can reduce traffic congestion and allow the network to operate at its optimum level. Traffic metering is one of the congestion management strategies that can be used to protect congested areas of urban networks from getting oversaturated. Metering signals, similar to those implemented on on-ramps, can be placed at the borders of the congested areas to regulate the number of vehicles. In our research, we formulated the traffic metering problem for urban networks as an optimization program. The objective of this program is to maximize the number of completed trips by optimizing the number of vehicles entering the congested areas. However, the program is computationally complex and cannot be solved with conventional optimization techniques. Thus, we proposed a customized solution technique so that the program can be solved within an optimality gap. Moreover, we developed a second solution technique so that the solutions can be found in real-time by compromising the optimality conditions for online applications.
Evaluation of Structural Contribution of Asphalt Mixtures Through Improved Viscoelastic-Based Distress Indicator Parameters
RASOOL NEMATI, University of New HampshireShow Abstract
New Hampshire Department of Transportation (NHDOT) uses
the AASHTO empirical method for structural design of highways. This method uses
material layer coefficients to quantify the pavement structural capacity. The
coefficients used by NHDOT are based on values suggested by AASHTO in 1960s.
These coefficients are experimentally developed values and almost no fundamental
mixture property in their original development. On the other hand, the
evolutions in the mixture design and the growing traffic demand etc. requires
the reevaluation of the existing layer coefficients.
Electric Vehicle Charging Infrastructure Design and Power Management Under Traffic Equilibrium
Roksana Asadi, SUNY College, Stony BrookShow Abstract
Our Research regarding as “Joint Optimization of
Electric Vehicle Charging Infrastructure and Power Distribution Network Design
under Traffic Equilibrium” presents an integrated electric vehicle (EV) charging
infrastructure and power distribution network (PDN) design in a bi-level model
structure to capture optimal power flow under traffic equilibrium in urban
transportation networks. The upper level formulation aims to optimize the PDN’s
optimal energy dispatch and EV charging facility deployment under budget and EV
range limitation constraints. The lower level formulation seeks traffic
equilibrium, where EV users aim to minimize their travel time. The integrated
power distribution and transportation network design problem involves
non-linearity and mixed-integer variables. To reduce the complexity of the
problem, the proposed bi-level model is converted into an equivalent single-
level mixed-integer non-linear program (MINLP) using Karush-Kuhn-Tucker
conditions and the non-linear user and generators’ cost functions are
approximated by piece-wise linear functions, which yield a single-level
mixed-integer linear program (MILP). The problem is then solved using a column
and constraint generation methodology. We mathematically show the computational
efficiency and solution quality of the proposed algorithm compared to
Benders-dual technique. The proposed methodology is applied to two case study
networks of different sizes to evaluate its performance compared to benchmark
solutions. We have solved both single-level MINLP and MILP with deterministic
demand and compared their results to the proposed column and constraint
generation algorithm under uncertain demand. Furthermore, a series of
sensitivity analyses has been conducted to study the effect of input parameters
on the proposed methodology and draw managerial insights. The numerical
experiments confirm that the proposed algorithm can solve the problem
Distributed Optimization and Coordination Approach for Real-Time, Network-Level Traffic Signal Optimization in Connected Transportation Network
S.M.A. Bin Al Islam, Washington State UniversityShow Abstract
The primary focus of the research is to develop a
real-time traffic signal control strategy for connected urban-street networks.
This study proposed a distributed model predictive control (DMPC) that
decomposed a network-level signal timing optimization problem into several
intersection-level mixed integer linear optimization problem to ensure real-time
applicability. The proposed DMPC involves gathering connected vehicle (CV)
information at each intersection at each control step and optimizing the timing
of signalized intersections over a prediction horizon based on macroscopic
traffic flow dynamics. However, the first decisions on terminating or extending
the signals of the optimal control sequence is implemented for the next time
step. In addition, adjacent intersections share their optimal signal control
sequence to coordinate their decisions for upcoming prediction horizon and
ensure smooth traffic flow throughout the network. The algorithm maximized
intersection throughput by 1% to 13% and reduced travel time by 18% to 34%
increase compared to the optimized actuated coordinated signals under various
demand patterns in a simulated 100% CV environment.
Optimization of Road Weather Information System Sensor Location and Density Using Big Data
Simita Biswas, University of AlbertaShow Abstract
Prevention of weather-related road crashes is a vital and challenging issue, particularly for cold regions including Canada and the Unites States. Nearly 3,000 fatalities result from weather-related accidents in Canada and over 1.5 million road crashes – of which 800,000 lead to injuries and 7,000 fatalities - occur annually in the entire United States. Road Weather Information Systems (RWIS) are an innovative intelligent transportation system invention and play a critical role in improving safety and mobility of motorists by reducing weather-related crashes. RWIS consist of a series of environmental sensor stations (ESS) that are installed specifically to monitor and disseminate real-time road weather and surface conditions so that highway maintenance agencies can deliver more effective maintenance services and the traveling public can make better travel related decisions. Since RWIS are expensive to operate and maintain, it is of paramount importance to know the best number of stations and locations to maximize their spatiotemporal coverage and investment returns. Despite this significance, there have only been a few past efforts devoted to the RWIS location problem. In this research, geostatistical models and combinatorial algorithms are implemented to strategically plan and optimize RWIS network of U.S. states and Canadian provinces. A large-scale dataset; namely, remotely sensed satellite imagery, geographic information system data, and RWIS data from twenty states have been used to generate more conclusive RWIS sitting guidelines. It is anticipated that the developed guidelines can readily be used as a decision-making tool for RWIS network planning and expansion. This research further proposes to develop a maintenance decision support system (MDSS) that will benefit North American transportation authorities with improved safety, mobility, and productivity during inclement weather events.
Dynamic Pricing and Long-Term Planning Models for Managed Lanes with Multiple Entrances and Exits
Venktesh Pandey, University of Texas, AustinShow Abstract
Priced managed lanes are increasingly being considered as an alternative for improving travel time reliability on a corridor while generating revenue for infrastructure projects. In the recent years, networks of managed lanes have become increasingly complex having multiple entrances and exits that can span an entire corridor length across a city. This dissertation is motivated by three challenges with the current models for complex managed lane networks:
In this dissertation, we improve the managed lane models used for long-term planning and dynamic toll pricing by considering en route changes in the travelers’ route choice. We propose a route choice model called sub-route generation (SRG) model and present a methodological comparison with different route choice models. The SRG model generates an average percent error of 0.93% in the expected cost compared to the optimal route choice model based on Markov decision processes. We demonstrate the effectiveness of value function approximation (VFA) for dynamic pricing of managed lanes in both centralized and distributed settings. VFA generates 10-90% higher revenue and 0-27% lower total system travel time than the existing heuristics used in practice. We also develop efficient algorithms for solving multi-class user equilibrium with recourse for both static and dynamic settings for long-term planning which lead to an average 1.6% reduction in total system cost compared to equilibrium models without online information. These findings can improve existing state-of-the-art models for managed lanes that can lead to improved traffic and revenue forecasts, efficient use of real-time data for model calibration, and for improved online route guidance using navigation applications.
Stormwater Mitigation Through Fully Permeable Pavement
Yazan Al-Zubi, California State University, Long BeachShow Abstract
This project presents the implementation of a new design method for fully permeable pavements, developed using the mechanistic‐empirical approach by the University of California Pavement Research Center (UCPRC) through building two test sections at California State University Long Beach (CSULB).
Fully permeable pavements are characterized as that in which all layers are permeable, and the pavement structure serves as a reservoir to store water and minimize the negative impacts of stormwater spillover. The California Department of Transportation (Caltrans) has shown interest in developing fully permeable pavement design for use in regions that convey substantial truck activity as a potential stormwater management, utilizing best management practice to give low‐effect infrastructure and proficient framework operation.
A location was selected within CSULB for the construction of the test sections. Pressure cells and strain gages were installed during the construction of pavement for measuring the stress on top of the subgrade and strain at the bottom of surface layer on both test sections to assess the performance of the fully permeable pavement. The data from pressure cells and strain gages were analyzed using MATLAB program and graphs were plotted to study the pattern in the data sets. The plots revealed that the asphalt section experienced more stress and strain in comparison to concrete test section. The traffic count was also determined.
Both test sections showed reliable performance in terms of distresses, and performed well in terms of infiltration during 2017 ‐ one of the wettest years in California. Based on the performance evaluation of both test sections, the fully permeable pavement design will be enhanced and developed as a potential best management practice for stormwater mitigation.
Robust Analytical Chip Seal Design Method Using 3D Laser Scanner
Yorguo El Hachem, University of Texas, AustinShow Abstract
People pay taxes and expect to drive on smooth safe
roads, but pavements are designed to fail; what an unfortunate shortcoming! Chip
seal is a popular and cost-effective pavement preventative treatment used
worldwide by highway agencies and state DOTs.