Graduate Research Symposium

Welcome

The Civil & Environmental Engineering (CEE) Department is pleased to present the 2021 Graduate Research Symposium. The first symposium was held in 2019 in person. This year, our online format will be hosted via Zoom on April 9, 2021. The program will start at 4PM Eastern.

Six CEE students have submitted their research presentations (preview videos below). The program will start with opening remarks by CEE Chair Dr. John Daniels and symposium coordinator Dr. Olya Keen. While judges score the presentations, symposium guests can virtually visit student presenters via Zoom rooms. The symposium program will last approximately 1 hour. The event will wrap up with the announcement of the three winners. Best wishes, everyone!

Research Submissions

1. SUBMISSION A
   • Title: Impacts of CAVs on Transfer-based DQN Controlled Signal Intersection: Insights from Mixed Traffic and Information Levels
     • Presenter: Li Song
     • Committee Chair: Dr. Wei Fan
     • Synopsis: With the Vehicle to Infrastructure (V2I) technology, the signal controller of the intersection could be more intelligent based on the information provided by the connected and automated vehicles (CAVs). A deep reinforcement learning approach is proposed for signal controllers. As the direct training procedure of the deep reinforcement learning would be cumbersome and unstable, this study improved the training procedure of the Deep Q Network (DQN) by transferring the prior information from previous models with similar traffic scenarios. Different traffic demands (from low, medium, medium-high, to high) and model settings are tested. Results indicate that the prior information provided by the Transfer based DQN could improve the training efficiency and stability. Compared to, controlled signal. Meanwhile, the results of 100% market penetration rates (MPRs) of CAVs indicate that the transfer-based DQN approach could improve the system performance (total waiting time, average stopped vehicle, CO2, and fuel consumption) compared to fixed signal schemes. Also, for scenarios with different MPRs of CAVs (from 100% to 20%), 20% MPRs of CAVs (20% information levels of total vehicles) show a worse system performance compared to fixed signal schemes. Moreover, the system performance is improved after 40% MPRs of CAVs. This indicates that the DQN controlled signal intersection system requires a certain vehicle information level between 20% and 40%. The insights of this study should be valuable to design intelligent signal intersections, improve training efficiency and performance, and provide guidance for field experiments of the DQN signal intersection systems.
     • Video Presentation Link: Click to view Submission A
2. SUBMISSION B
   • Title: The Impact of Connected and Autonomous Vehicles on the Performance of Superstreet
     • Presenter: Shaojie Liu
     • Committee Chair: Dr. Wei Fan
     • Synopsis: This presentation introduces the research work investigating connected and autonomous vehicles (CAVs) in the environment of superstreet. A mixed traffic environment where both human-driven traffic and CAV traffic were considered and modeled by different car following models. CAVs was assumed to connected to signal status and adjust speeds accordingly. A real-world superstreet was selected for a case study and the car following model for human driven traffic was calibrated with genetic algorithm. The simulation results demonstrated the superiority of CAV in the environment of superstreet and average speed were improved by about 10 percentage in peak hour. Different improvements were observed for main street and cross street, which may be explained by different traffic volume patterns present on the major street and cross street.
     • Video Presentation Link: Click to view Submission B
3. SUBMISSION C
   • Title: Hydroxyl Radical and Sulphate Radical as Treatment Strategies for Antibiotic Resistance Control
     • Presenter: Adeola Sorinolu
     • Committee Chair: Dr. Mariya Munir
     • Synopsis: Nowadays, treatable illnesses are often fatal due to the dwindling effectiveness of antimicrobial drugs. Antibiotic-resistant bacteria (ARB) and antibiotic resistance genes (ARGs) are the indicators of antibiotic resistance (AR) in any environment. The prevalence of ARB and ARGs in the environment is linked to the overuse of antibiotics in human and animal treatment. Excess antibiotics discarded as well as antibiotics in excrement end up in the wastewater treatment plants (WWTPs). Consequently, WWTPs are identified as reservoirs and sources for the release of ARB and ARGs into water sources. The effective removal of ARB and ARGs from wastewater by disinfection before discharge into receiving water bodies will help mitigate the spread of AR through water. Advanced oxidation processes (AOPs) that utilize strong oxidizing power of hydroxyl radical ({OH}^.) and sulphate radical (S{O_4}^{.-}) have received growing interest. AOPs have been successfully used to eliminate recalcitrant micropollutants. Hence, AOPs are promising technologies for ARGs degradation. Several studies exist on the removal of ARB and ARGs from wastewater using both conventional disinfection treatments and AOPs. However, very few have compared the kinetics of extracellular and intracellular ARGs degradation using {OH}^. and S{O_4}^{.-}. Also, limited studies have investigated the biological activities of the ARGs downstream of the disinfection process. An understanding of these fundamental issues is essential to the successful control of AR by AOPs. The goal of my research is to obtain results that will help optimize AOPs to ensure effective removal of these AR determinants from wastewater.
     • Video Presentation Link: Click to view Submission C
4. SUBMISSION D
   • Title: Surveillance of COVID-19 Outbreak in Local Community Through Quantitative Analysis of SARS-CoV-2 in Wastewater
     • Presenter: Md Ariful Islam Juel
     • Committee Chair: Dr. Mariya Munir
     • Synopsis: Wastewater base epidemiology (WBE), a new line of research that has already drawn a significant attention to the environmental science and engineering academic community, can be implement as an early warning tool using COVID-19 wastewater surveillance data and public health data. However, precise and accurate viral copies quantification in wastewater and data normalization prior to correlating with COVID-19 cases which are pre-requisite for making that tool successful. The purpose of this paper is to quantify SARS-CoV-2 concentration in wastewater and correlating this data with local COVID-19 cases in purpose of implementing WBE for preventing community level outbreak in Charlotte area. Electronegative membrane filtration (EMF) has been used for virus concentration and Nucleocapsid gene (N1 and N2) based primers and probe set used for the SARS-CoV-2 RNA viruses quantification through RT-qPCR. A known concentration of the Bovine Corona Virus, a surrogate of Human corona virus, was spiked in the wastewater for the overall process control of the system. Most of the samples showed the prevalence of SARS-CoV-2 viruses in wastewater in the range of 1000 to 1000000 copies/L which was resulted due to the recent COVID-19 case surges in the community. An average of 8 to 15 % of Bovine Coronavirus recovered as an overall process control while 40 – 50% hepatitis G virus recovered from the RNA extraction process which implies that a significant portion of viruses were lost during the sample processing stage and that can be considered for the normalization SARS-CoV-2 quantification in wastewater.
     • Video Presentation Link: Click to view Submission D

5. SUBMISSION E
   • Title: Engineered Water Repellency for Frost Heave Mitigation
     • Presenter: Micheal Uduebor
     • Committee Chair: Dr. John Daniels
     • Synopsis: Cold weather and frost action have a major effect on the design, construction, performance, and maintenance of roadways. This includes any area which experiences seasonally cold weather, including North Carolina. Frost heaving and thaw weakening are especially problematic, subjecting all elements of a pavement system to significant changes in moisture content, stress, and strain. Nationally, this leads to recurrent annual maintenance costs estimated at over 2 billion dollars, as well as additional economic impacts because of related vehicle damage, road closures, and weight restrictions. Studies identify three basic requirements for frost action; freezing temperatures, availability of water, and frost-susceptible soils. In North Carolina, this can be especially relevant in roads along north facing slopes in the Blue Ridge mountains region. While advances have been made designing for freezing temperatures and providing for groundwater separation, very little progress has been made in terms of in situ soil improvement. A cost and labor-intensive approach is to undercut and replace unsuitable soils. As an alternative, this presentation will describe Engineered Water Repellency (EWR), a process in which soils are made hydrophobic. This is achieved by combining soils with cost-effective and environmentally compatible polymers and other complex organic molecules. This process is being explored through a multi-year project funded by the U.S. National Science foundation, including laboratory, field, and numerical studies. This presentation will provide an overview of the study and results obtained thus far.
     • Video Presentation Link: Click to view Submission E
6. SUBMISSION F
   • Title: Travel Time Forecasting on a Freeway Corridor: A Dynamic Information Fusion Model Based on the Random Forests Approach
     • Presenter: Bo Qiu
     • Committee Chair: Dr. Wei Fan
     • Synopsis: Recently, the need for travel time prediction has become indispensable due to the increasing congestion in the roadway network. However, travel time prediction is highly complex as it is affected by a wide variety of factors. The acquisition and popularization of big data in the field of transportation have enabled the collection and diffusion of real-time traffic information. Different machine learning approaches have been employed by different researchers, and the results indicate that such approaches can give better performances than traditional models. However, such machine learning methods are practically faced with an overfitting problem that is difficult to overcome. Especially, when the traffic conditions greatly change, the prediction results are often unsatisfactory. In addition, the random forests (RF) method has a very good Bias-Variance trade-off which can help avoid the overfitting problem. This research develops an RF method to predict the freeway travel time by using the probe vehicle-based traffic data, and therefore helps to gain a better understanding of how different contributing factors might affect travel time on freeways. Detailed information about the input variables and data pre-processing is presented. Findings Parameters of the RF model are estimated, the parameter tuning process is also discussed, and the relative importance of each variable and their ranks in the RF model are also presented. Results indicate that RF always produces more accurate travel time prediction.
     • Video Presentation Link: Click to view Submission F

Questions/Comments

Please forward your logistics questions or comments to Linda Hargrove (Linda.Hargrove@uncc.edu).

Thank you for participating in this symposium.