Our Research Projects

SHF: Small: Collaborative Research: Accelerated Data Transformation: A Software-Hardware Stack for Transducers

Michela Becchi
10/01/2019 - 09/30/2022

Big Data’s growing importance is evident from broad application to business, public policy, medicine, and research. Many Big Data applications perform frequent data transformations on unstructured data. Data transformations can be mapped onto finite state transducers – a computational model with a solid theoretical foundation. The goal of this work is to design and develop a software stack for transducer processing that supports diverse platforms such as CPU’s, GPU’s, and efficient data-intensive accelerators (such as our Unstructured Data Processor). We summarize our research efforts as follows. First, we will create a high-level interface consisting of sets of production rules that can be mapped on transducers and can support a variety of data transformation operations. This interface will enable specification of flexible, extensible, and composable transducer programs. Second, we will build a sophisticated compiler that maps the high-level programming interface onto the finite state transducer computational model and includes optimization techniques that exploit the properties of this model. This compiler will produce an intermediate representation that can be mapped onto diverse hardware. Third, we will address software challenges involved with mapping the optimized finite state transducers onto a data-intensive accelerator, managing data-parallelism, limited memory, and generation of compact and efficient code that leverages the hardware features of the accelerator, particularly specialized operations. Finally, we will investigate the limitations of the finite state transducer model and extend it so to support more general data transformations performed in modern data analytics system. An example is block-based data compression and decompression, which we have demontrated can be efficiently supported by data-intensive accelerators, but cannot be expressed in the transducer model. We call this model unbounded transducers.

Green Energy Hub – DGI-IEM Demonstration Industry, FREEDM Core Project

David Lee Lubkeman, Ning Lu, & Mesut E. Baran
08/15/2010 - 06/30/2020

The Green Energy Hub testbed is an integrated hardware system demonstration incorporating technologies from the Enabling Technology and Fundamental Science research planes.

ARO IPA: Program Manager, Processing and Fusion (Electronics Engineer)

Hamid Krim
06/24/2019 - 06/23/2021

An interest in a position at ARO in the capacity of an IPA contract employee is expressed. The interest is particularly focussed in exploiting the PI’s knowledge and familiarity with DOD challenges to provide the vision, leadership and help shape the research agenda priorities to ensure that the current as well future sensing and information and machine learning challenges be met.

Feeder Anti-Islanding Detection Using HIL Modelling and Simulation

Mesut E. Baran, Srdjan Miodrag Lukic
05/15/2019 - 09/30/2019

This project will focus on hardware in loop analysis of faults within the feeder that uses the MATLAB/Simulink model of Mini D-VAR, Opal-RT software for HIL simulations.

The study will include using a feeder model and simulating different types of faults on the feeder and determining if for any of these faults, the protection and control logic implemented inside relays can potentially detect the fault, therefore not causing the creation of an island.

PD-04: Development and Demonstration of a Power Electronics Assisted Distribution Voltage Regulator

Mesut E. Baran
01/02/2019 - 12/31/2019

This project aims at developing a novel power electronics assisted voltage regulator for the distribution grid. The proposed solution is based on the existing Step Voltage Regulator (SVR) but potentially increases their number of electrical operations by five times. The proposed solution offers arcless operation, fast dynamic response, volt/var control, and voltage sag/swell compensation capability.

Learning Novel Feature Representations from Deep Neural Networks (Image Deep Structure Learning and Inference Exploration)

Hamid Krim
03/21/2019 - 03/31/2021

This work is to thoroughly evaluate and test algorithms that have been developed within VISSTA Laboratory. These pursued remote sensing data sets with associated applications will strengthen our collaborative work with Lawrence Livermore by a number of internship periods of the student at LLNL.

2019 International RoboSub Competition

John F. Muth
11/15/2018 - 08/31/2019

The underwater robotics club competes in the AUVSI RoboSub competition, an international robotics competition sponsored by the Association for Unmanned Vehicle Systems International and the US Office of Naval Research. In RoboSub, teams compete to create Autonomous Underwater Vehicles (AUVs) that can navigate an obstacle course and complete tasks underwater with no human input whatsoever. The tasks are designed to be similar to the real world, such as searching for and retrieving objects on the floor of the pool, manipulating levers and wheels, avoiding obstacles, and locating an acoustic “pinger” (sonar transmitter), much like the ones that are used to locate the “black box” on downed airplanes. Every year, students from roughly 40 teams around the world decide to take this challenge upon themselves, and learn real engineering, problem-solving, and teamwork skills along the way.

Real-Time Remaining Useful Life Assessment for Batteries using the SBG at Butler Farm

Mo-Yuen Chow
05/16/2019 - 08/15/2019

The energy storage systems at Butler Farms are currently being monitored and maintained by NCEMC using the Smart Battery Gauge developed during the Smart Battery Gauge for Continuous Battery Assessment at Butler Farm and Smart Battery Gauge for Continuous Battery Health Assessment at Butler Farm projects. The Smart Battery Gauge was developed to continuously monitor and provide live feedback about the State of Charge and State of Health of the energy storage system at a rack level. This project will develop a Remaining Useful Life (RUL) Assessment algorithm to assist in determining State of Function (SOF).

Fusion Algorithm Development of Multi-Modality Data

Hamid Krim
02/25/2019 - 12/31/2021

Our goal in this effort is to apply recently developed multi-modality fusion algorithms to vehicle data collected at ORNL. The testing of additional sensors would ensure the applicability of our approaches to the ORNL target goal of potentially deploying these algorithms in their system which is an on-going project.
This testing project is to hence harness a number of sensing modalities including, acoustic, magnetic, video, possibly laser, and jointly exploit them to carry out target inference. We will focus on a vast database of vehicles most of which have been characterized by the previously mentioned modalities.
In coordination with the ORNL Project Lead, we will work to identify the potential stress factors for the developed algorithms in the field, and will define the proper strategies to preserve a robust or at least gracefully degrading performance of the algorithms.

Evaluation of Novel Approaches for Early Detection of and Detailed Characterization of Maize Foliar Disease

Peter Ojiambo, Peter J. Balint-Kurti, & Michael Kudenov
02/01/2019 - 01/31/2020

We generally assess levels of foliar disease in corn by observation in the field using a visual scale. While this method is robust and gives reproducible data, it does not provide qualitative data such as lesion size or shape, nor does it give us good data on speed of disease progression or on timing of initial symptoms. We will rate a set of 50 genetically similar lines each of which has a different allele conferring resistance to the foliar disease southern leaf blight. We will use a variety of approaches to rate disease , including hyperspectral imaging and digital imaging. We will also rate them conventionally at very frequent intervals. Finally, we will measure yield.

This work will help to develop methods for the early detection of foliar disease in the field. It will also characterize disease progression and the relationship between symptoms and yield loss. This knowledge will aid in the development of predictive models to guide farmers in the decision of whether and when to apply fungicides.

The characterization of different mechanisms used by different resistance genes will guide breeders in the combination of resistance genes to produce more optimally-robust disease resistant lines.

Enabling Side-Channel Attacks on Post-Quantum Protocols through Machine Learning, CAEML Core Project P18-13 funded with industry membership dues

Aydin Aysu
01/01/2019 - 12/31/2019

This project is on using machine learning for trusted system design

A Prototype DNA Hard Drive

Albert J. Keung, James Tuck
07/01/2019 - 03/31/2020

This proposal seeks to construct a prototype information storage system using dna.

Fundamentals of Power Engineering to Support Integration of Distributed Energy Resources, CAPER Enhancement Project

Ning Lu, David Lee Lubkeman, & Mesut E. Baran
12/01/2018 - 08/31/2019

This project will develop course materials and instructor teaching aids to deliver a four-week course. The main goal of this course is to provide newly graduated engineering students and power professionals with a
quick but broad introduction on power engineering fundamentals related with the integration of distributed energy resources (DERs). This course materials will include fundamentals of power system design and operations pertinent to the adoption of DERs (e.g. energy storage, solar photovoltaics, controllable loads,
etc.) in power system operation and planning. Many engineering graduates today did not receive the necessary training on the basics of three-phase electric power because most universities no longer require power systems be taken by all electrical engineering students. This course will cover basic topics usually
taught in a two-semester elective power engineering curriculum. After taking this condensed course, engineers with a general engineering background will be equipped to master the following fundamental materials on how to operate and plan a three-phase electricity power system: the modeling of power
system components, power flow studies, economic dispatch, unit commitment, power system dynamic response, the modeling of microgrid and DERs.

High-Dimensional Structural Inference for Non-Linear Deep Markov or State Space Time Series Models, Center for Advanced Electronics through Machine Learning (CAEML) Core Project 3A4

Dror Zeev Baron, William R. Davis, & Paul D. Franzon
01/01/2019 - 12/31/2019

The project will explore deep Markov models for high dimensional time series. While past works on density estimation for multi-dimensional latent time series systems have focused on low- to medium- dimensional settings, we will try to move to higher dimensional settings.

Fast, Accurate PPA Model-Extraction, Center for Advanced Electronics through Machine Learning (CAEML) Core Project 3A2 funded with industry membership dues.

William R. Davis, Paul D. Franzon, & Dror Zeev Baron
01/01/2019 - 12/31/2020

This project researches methods for extracting fast and accurate estimators from System-Level Architecture to Power, performance, and area (SLA2PPA) in digital integrated circuits. Specifically, this project focuses on elimination of the complicated gate-level simulations needed to make accurate predictions of power, which typically occur very late in the design process. Extraction of system-level power models is extremely difficult, because the data-points are so few and so noisy, while the number of possible model parameters is so huge. This project will develop a comprehensive data-mining methodology to maximize the accuracy of PPA predictions while minimizing the data-collection effort.

FREEDM Electric Machines and Drives Research, FREEDM Core Project

Iqbal Husain
07/01/2018 - 06/30/2020

Electric machines and drives are a central area of research for FREEDM and are supported by our industry members.

High Precision Measurements of Beta Decay Using Stored Ultracold Neutrons and Cold Neutron Beams

Albert R. Young, Paul R. Huffman, & Daniel D Stancil
07/15/2019 - 06/30/2022

This proposal covers measurements of neutron beta decay using a cold neutron beam at the Spallation Neutron Source and ultracold neutron (UCN) experiments at the Los Alamos National laboratory. The goal of these experiments are measurements which reduce the reduce the current uncertainties in the neutron lifetime to the level of about 1 second (the lifetime is 880 seconds in the latest average of the global data set) and the uncertainty in the axial coupling constant to about 0.025%. These quantities are important for high precision predictions of the energy output of the sun, the cross-sections for neutrinos coming from reactors and searches for physics beyond the standard model of particle physics.

Each experiment has unique strengths, in that Nab uses a cold neutron beam to provide an intense, localized source of neutrons, ideal for a time-of-flight spectrometer. The other two experiments, UCNA and UCNtau, use UCNs (neutrons going so slowly they can be stored in material vessels and magnetic traps) to help overcome systematic errors. The UCN measurements are stationed at the LANL UCN source, a solid deuterium source developed through earlier efforts co-led by our NCSU group.

Evaluation of Sensor Technology for Real-Time Detection of Cyst Nematode Infected Soybean Plants

Ralph A. Dean, Omer Oralkan
12/01/2018 - 11/05/2019

Soybean cyst nematode (SCN, Heterodera glycines) is a major pest in all soybean-growing areas, particularly in sandy soils. Here, we propose a series of growth chamber experiments to demonstrate the potential of sensor technologies based on the capacitive micromachined ultrasonic transducer (CMUT) developed by our research team at NC State for detecting volatiles associated with SCN. In addition, we plan to identify and characterize discriminating VOCs from SCN infected soybeans and examine their detection by the sensor arrays.

Adopting D-VAR to Mitigate PV Impacts on a Distribution System

Mesut E. Baran, Srdjan Miodrag Lukic
01/01/2019 - 12/31/2019

D-VAR is an emerging technology which is designed for distribution level volt/VAR applications. The goal in the first phase of the study is to quantify the benefits a D-VAR will offer for a given distribution feeder with high penetration of PV. The system that will be considered is a Duke feeder with large PV. Hence, the main focus of this study is to assess how effective mini D-VAR will be in mitigating these impacts on various distribution feeders in Duke Energy service territory.

The study will first simulate a given feeder to investigate the PV impacts on the circuit and then focus on the use of D-VAR to mitigate these issues.

Three-Phase T-Type Inverter

Subhashish Bhattacharya
11/06/2018 - 12/09/2019

Initially introduced to mitigate losses in medium voltage drives, the development of Wide Bandgap (WBG) semiconductors have now made multilevel inverters a viable alternative to the conventional two level inverter at low voltages also. The benefits of multilevel inverters over two-level inverters are given below:
1. Reduction of voltage transients at the motor windings
2. Lower harmonic distortion in the output current and voltage
3. Reduction in output filter size
4. Lower common mode currents
5. Reduced stresses on the power switches
The Three-Phase T-Type Multilevel Inverter in specific offers superior performance, with the conduction losses of a two-level inverter and the switching losses of a three-level inverter. These advantages allow the T-Type inverter to be operated at very high switching frequencies, making them ideal for high-speed motor drives. Also, the lower THD in output current produces less torque ripple, making them suitable for high precision servo drives.