In large-scale power systems, the integration of intelligent monitoring system increases the system resiliency and the control robustness. For example, sensor monitoring allows to automatically supervise the health of sensors and detect sensor failures without relying on hardware redundancy, and hence, it will further reduce the cost of monitoring systems in power systems. Sensor failure is critical in smart grids, where controllers rely on healthy measurements from different sensors to determine all kinds of operations. Current literature review shows that most of the researchers focus on control and management side of smart grids, assuming the information control centers or agencies get from sensors is accurate. However, when sensor failure happens, missing data and/or bad data can flow into control and management systems, which may lead to potential malfunction or even power system failures. This brings the need for Sensor Fault Detection and Isolation (SFDI), to eliminate this potential threat. The integration of the SFDI into monitoring systems will allow avoiding the contingencies due to fault data, and therefore increases the system resiliency and the control robustness. Hardware redundancy is the common solution for SFDI. By placing multiple sensors in the same position, the control center can then rely on redundant sensors when one is broken or inaccurate. Apparently, this method will increase the cost significantly when applying to large power systems. Analytical redundancy, on the contrary, a quantitative method built from power system models, is a more promising solution. It does not necessarily require hardware redundancy and hence can lower the cost. With an appropriate number of sensors placed in strategic locations, the algorithm can then automatically detect sensor failures without the need of extra redundant sensors. Furthermore, SFDI together with intelligent sensor optimization and placement will also facilitate the transfer of conventional central grid control to distributed decision making agencies with minimum computation and communication burden for each branch, and thus, it will enhance the system performance and resiliency. In this dissertation, a comprehensive review over the state-of-the-art FDI methodologies is given at first, then a proposed algorithm to determine the optimal location of computation agents is introduced, which serves as a guide for the SFDI algorithm implementation explained right after. The results of the algorithms indicated promising application in power system monitoring. / A Dissertation submitted to the Department of Electrical and Computer Engineering in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Summer Semester 2018. / July 19, 2018. / communication cost, computation agent, computation cost, distributed computation, optimal location, sensor fault / Includes bibliographical references. / Chris S. Edrington, Professor Directing Dissertation; Juan C. Ordóñez, University Representative; Pedro Moss, Committee Member; Simon Foo, Committee Member.

Nowadays the medium-voltage dc system (MVDC) has been proposed in the renewable energy collector fields, long distance power transmission, small-scale industrial networks and all-electric shipboards due to its relatively higher efficiency, higher flexibility and lower cost in certain applications compared to the ac grid. Batteries offer scalable energy storage solutions in these applications for high-power and long-term energy demands with high energy density. Batterers play an essential role to smooth the power fluctuations and stabilize the grid as well. As the interface between battery energy storage and MVDC bus, the battery energy storage system (BESS) converter is a key enabling technology with specific requirements. Due to the lack of mature dc circuit breakers, the BESS converter is desired to achieve superior dc fault response which benefits the MVDC system reliability and resiliency. In addition, considering the high expenses and limited lifetime of nowadays battery products, multiple services and functions are preferred for BESS. In this research, the isolated modular multilevel dc/dc converter (iM2DC) based BESS is proposed. It can achieve both fault current limiting and fault ride through functions with direct dc current control capability, so it is possible to maintain the system operation during fault to ensure fault localization and fast recovery. Besides, via the virtual impedance method, the proposed topology employs the converter cell capacitors rather than batteries to provide the ripple energy to achieve the active power filter (APF) function, which allows the energy storage system to improve MVDC system power quality without consuming battery lifetime or extra circuits. In addition, since the medium-frequency transformer operation frequency can be as high as the converter switching frequency, the whole system power density will be improved. A controller hardware-in-the-loop testbed, which consists of the iM2DC based BESS model simulated in the real-time digital simulator (RTDS) and the multifunctional control programmed in the ABB controller products, is utilized to validate the functionality of proposed technology. Furthermore, the system efficiency of proposed BESS is not most optimized with the sinusoidal modulation. Therefore, in this research, a novel phase-shifted square wave modulation strategy is proposed for iM2DC. Compared to the conventional modulation methods, the proposed technique achieves reduced dc inductance due to higher equivalent switching frequency. In addition, the required capacitor energy can be minimized, which decreases the capacitor size without sacrificing the total device rating. Detailed principles of the proposed modulation and passive components design are presented. A downscaled 2kW prototype is built in the lab and the experimental results are provided to demonstrate the proposed modulation strategy. Finally the dissertation work is summarized and the scope of future work is discussed. / A Dissertation submitted to the Department of Electrical and Computer Engineering in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Fall Semester 2017. / November 17, 2017. / battery energy storage system, high/medium voltage dc grid, isolated modular multilevel dc/dc converter / Includes bibliographical references. / Hui Li, Professor Directing Dissertation; Juan C. Ordonez, University Representative; Thomas A. Lipo, Committee Member; Chris S. Edrington, Committee Member; Michael Steurer, Committee Member.

In undersea remote sensing change detection is the process of detecting changes from pairs of multi-temporal sonar images of the seafloor that are surveyed approximately from the same location. The problem of change detection, subsequent anomaly feature extraction, and false alarms reduction is complicated due to several factors such as the presence of random speckle pattern in the images, variability in the seafloor environmental conditions, and platform instabilities. These complications make the detection and classification of targets difficult. This thesis presents the first successful development of an end-to-end automated seabed change detection using multi-temporal synthetic aperture sonar (SAS) imagery that include a false detection/false alarms reduction based on principal and independent component analysis (PCA/ICA). ICA is a well-established statistical signal processing technique that aims to decompose a set of multivariate signals, i.e., SAS images, into a basis of statistically independent data-vectors with minimal loss of information content. The goal of ICA is to linearly transform the data such that the transformed variables are as statistically independent from each other as possible. The changes in the scene are detected in reduced second or higher order dependencies by ICA. Thus removing dependencies will leave the change features that will be further analyzed for detection and classification. Test results of the proposed method on a data set of SAS images (snippets) of declared changes from an automated change detection (ACD) process will be presented. These results illustrate the effectiveness of component analysis for reduction of false alarms in ACD process. In the context of sea surface object detection, this thesis investigates bistatic radar engagement using synthetic aperture radar (SAR) and examines five models of the bistatic electromagnetic scattering that will support future research on SAR sea-surface change detection. / A Dissertation submitted to the Department of Electrical and Computer Engineering in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Fall Semester 2017. / November 21, 2017. / ACD, CCD, Change detection, Electromagnetic scattering, ICA, synthetic aperture sonar / Includes bibliographical references. / Rodney G. Roberts, Professor Directing Dissertation; Anke Meyer-Baese, University Representative; Uwe H. Meyer-Baese, Committee Member; Simon Y. Foo, Committee Member.

Power hardware-in-the-loop (PHIL) simulation is a technique whereby actual power hardware is interfaced to a virtual surrounding system through PHIL interfaces making use of power amplifiers and/or actuators. PHIL simulation is often an attractive approach for early integration testing of devices, allowing testing with unrealized systems with substantial flexibility. However, while PHIL simulation offers a number of potential benefits, there are also a number of associated challenges and limitations stemming from the non-ideal aspects of the PHIL interface. These can affect the accuracy of the experiments and, in some cases, lead to instabilities. Consequently, accuracy, stability, and sensitivity to disturbances are some of the most important considerations in the analysis and design of PHIL simulation experiments, and the development of PHIL interface algorithms (IA) and augmentations for improvements in these areas is the subject of active research. Another area of research sharing some common attributes with PHIL simulation is the field of robotic bilateral teleoperation systems. While there are some distinctions and differences in characteristics between the two fields, much of the literature is also focused on the development of algorithms and techniques for coupling objects. A number of disparate algorithms and augmentations have also been proposed in the teleoperation literature, some of which are fundamentally very similar to those applied in PHIL simulation. While some of the teleoperation methods may have limited applicability in PHIL experiments, others have substantial relevance and may lend themselves to improvements in the PHIL application area. This work focuses on the application and analysis of a teleoperation framework in the context of PHIL simulation. The extended Lawrence Architecture (ELA) is a framework unifying and describing a large set of teleoperation interfacing algorithms. This work focuses on the application and analysis of the ELA to PHIL simulation. This includes the expression of existing PHIL IAs in the context of the ELA, derivation of relevant transfer functions and metrics for assessment of performance, the exploration of the implications of the transparency requirements, and the exploration of new IAs supported by the ELA which may be well suited to the particular characteristics of PHIL simulation. / A Dissertation submitted to the Department of Electrical and Computer Engineering in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Spring Semester 2018. / March 9, 2018. / Power Systems, Simulation / Includes bibliographical references. / Chris S. Edrington, Professor Directing Dissertation; Omer Arda Vanli, University Representative; Michael Steurer, Committee Member; Rodney G. Roberts, Committee Member; Md Omar Faruque, Committee Member.

Vehicular communication network consists of different wireless communication technologies working in conjunction with each other. These different wireless communication technologies have different technical parameters. Wireless communication technology includes Dedicated Short-Range Communication, WiFi, WiMAX etc. depending upon their network range, data bit transfer rate, safe effective maximum intended communication range, modulation technique adopted and many more, are deployed for specific safety application. The main objective of Intelligent Transport System (ITS) is Safety. Under safety application there are many objectives including safe approach to the intersection, pre and post-crash warning, total loss control correction etc. these safety applications require specific parameter of communication technology i.e. for safe intersection approach data bit rate need not to be high and other safety application seeks different parameter. It is obvious that no single wireless communication technology could fulfill all the specifics of communication technology and objective of ITS. In this research important wireless communication discussed. Their pros and cons are summarized in the vehicular environment. In order to show the importance of wireless communication technology in Vehicular network, one among many safety applications is simulated. In the simulation, safe approach to unsignalized intersection is simulated. Simulation is performed on VISSIM software developed by PTV group, Germany. Simulation is based on Nakagami Wireless probabilistic model under relaxed radio condition (no interferences) and finally conclusion is made. / A Thesis submitted to the Department of Electrical and Computer Engineering in partial fulfillment of the requirements for the degree of Master of Science. / Fall Semester 2018. / November 5, 2018. / Includes bibliographical references. / Bruce Harvey, Professor Directing Thesis; Simon Foo, Committee Member; Shonda Bernadin, Committee Member.

The modern day smart grid technology relies heavily on data acquisition and analysis. A distributed controller governs smart microgrid functions with one or more renewable sources and smart controllable loads. This sort of intelligent, scalable system is the primary drive for the Energy Internet (EI). Hence, in modern-day power systems engineering to analyze, understand and make efficient system design choices that capture robustness and scalability, Hardware in the Loop (HIL) simulations are required. Real-Time Simulations (RTS) is the state of the art technology thrusting the capstone of innovation for this industry. As engineers, we can model, simulate and validate smart grids operations more rapidly, robustly and reliably using RTS. With enough smaller time step for the simulation, the boundary between the real and the simulated systems slowly vanishes. It also enables the system to be simulated as Controller Hardware in the Loop (CHIL) or Power Hardware in the Loop (PHIL) setups, evolving and imitating the real physical world. The HIL (Hardware in the Loop) setup also enables a real data source or sink to be in the system to form the loop of exchange between the simulated system and real-world hardware which is most often a control hardware. The implementation of such a setup is made possible at Center for Advanced Power Systems (CAPS), named as Hardware in the Loop Test-Bed (HIL-TB). This evaluation architecture provides a systematic solution to HIL simulations. Now the sampling time for real-world sensors is generally in the order of microseconds, enabling this collected data to emulate the cyber-physical domain accurately. Thus, the challenge previously was to address the throughput of real-world input data into the simulated system efficiently and correctly. The quality of the Design of Simulation (DoS) using the real world data in the form of Real Time Input Profile (RTIP), improves, affects the quality of response of the real-time cyber-physical system simulation. Thus great care needs to be taken to prepare, prune and project the RTIPs to improve and enhance the system performance evaluation index. To solve this problem, partially successful attempts have been made in the direction of machine learning by using methods like clustering and regression to characterize large input profiles or by breaking them into subsections using fixed length sliding window techniques. These classic methods then perform data analysis on those sub-pieces to distinguish among a variety of input profiles and assign an index. These sub-profiles or sections would be then loaded into the simulation as environmental input to represent the physical system in the HIL simulations. This traditional procedure is observed to be arbitrary because clustering algorithms and metrics for methods like regression or classification are user-defined and there exists no standard practice to deal with huge input profiles. There have also been confusions regarding the size of the sliding window to create subsections, subsection joining logic, etc. Thus, to address this issue, the primary focus of this study is to present a systematic, controlled, reliable procedure to explore, screen, crop large input profiles and then to compress the same by selecting sections with most relative importance using a modified version of “knapsack” dynamic programming algorithm. This compression primarily aims to shrink down the total simulation time without much loss of information. The latter part of this study focuses towards response driven performance evaluation of the HIL simulations. This is ensured by targeted compression of original input profile based on the certain requirement of the simulation. This approach ensures that the control algorithm (CHIL simulations) or any other system operator is driven in a specific direction in the simulation response space by effectively sampling the input parameters space. The fully automated HIL-TB evaluation framework aided with Input Profile Time Compression (IPTC) module delivers a fast-convergent validation for the performance evaluation with relatively similar system response. In this study, the IPTC module has been applied to seven load profiles to compress their temporal length by a third. The case study used for the simulation with these RTIPs is the Future Renewable Electric Energy Delivery and Management (FREEDM) IEEE seven node system. The test results show great coherence between the uncompressed and compressed response and validate the performance of the IPTC module applied to real-world HIL simulations. Thus, it can conclude that the functionality of the IPTC module is validated by the quality of simulation response gained out of the compressed simulation as compared to uncompressed simulation. In future, endeavors can be made in this path by expanding the functionality of this compression module to not only identifying and managing important sections based on some initial assumption about the objective of the control application but also providing cognitive, autonomous understanding of the behavior of the controls and using that knowledge accomplishing compression of large input profiles. / A Thesis submitted to the Department of Electrical and Computer Engineering in partial fulfillment of the requirements for the degree of Master of Science. / Fall Semester 2017. / November 15, 2017. / Design of Simulation, Hardware in the Loop Simulation, Input Profile Compression, Real Time Simulation, Time Compression / Includes bibliographical references. / Omar Faruque, Professor Co-Directing Thesis; Mischa Steurer, Professor Co-Directing Thesis; Hui Li, Committee Member.

Digital Signal Processing (DSP) performs a very important role in various applications of electrical engineering like communications and signal enhancement. In many situations one finds that the DSP hardware available are fixed point processors. In these situations, it is necessary to perform DSP with high accuracy using the least amount of hardware resources. This thesis looks into an approach to calculate the two dimensional Discrete Hirschman Transform (DHT), the inverse DHT, the Hirschman Cosine Transform (HCT) and the inverse HCT using fixed-point hardware. The complex coefficients required for the transform are calculated beforehand and saved as vectors. Special attention has been given to minimize errors due to scaling. The processed image and the original image does not show significant difference even for DFT or DCT length of 128. Mean square errors of -37 dB for the DHT and -40 dB for the HCT could be obtained for DFT and DCT lengths of 128. / A Thesis submitted to the Department of Electrical and Computer Engineering in partial fulfillment of the requirements for the degree of Master of Science. / Fall Semester 2017. / November 17, 2017. / DCT, DFT, DHT, HCT, HOT / Includes bibliographical references. / Victor E. DeBrunner, Professor Directing Thesis; Linda DeBrunner, Committee Member; Bruce A. Harvey, Committee Member.

Power electronics technology has been widely used in electric power system to achieve high energy efficiency and high renewable energy penetration. Small signal instability phenomena could easily occur in systems with abundant power electronics because of high order passive elements and controller interactions among power converters. These instability issues degrade power quality or even cause system failure. Therefore it is necessary to build accurate small signal models for stability analysis and develop effective resonance mitigation techniques for stability improvement. The general stability analysis methods including eigenvalues-based method, component connection method, passivity-based method and impedance-based method have been evaluated and summarized. The impedance-based method is selected as the stability analysis tool for this research due to its advantages compared to other methods. Besides, three popular resonance suppression techniques, i.e. passive damper, active damper and virtual impedance control, are also studied and evaluated. The virtual impedance control is of interest because it does not reduce system efficiency or reliability compared to both the passive and active damper. A unified impedance-based stability criterion (UIBSC) has been proposed for paralleled grid-tied inverters. Compared to the traditional IBSC which evaluates all minor loop gains (MLGs) of individual inverter, the UIBSC assesses the derived global minor loop gain (GMLG) only once to determine system stability. As a result, the computation efforts can be significantly reduced when system contains a large number of inverters. In addition, a stability-oriented design guideline has been derived for the paralleled grid-tied inverters based on the GMLG. By using the guideline, the grid impedance, inverter filter parameters, time delays of digital control and control parameters can be analyzed or designed to meet the system stability requirement. The small signal stability of the FREEDM system is a critical issue due to the abundant power electronics devices and flexible control strategies. The impedance modeling methods for current controlled inverters, inverter stage of the SST, DAB converters are developed. The influences of control schemes on power converter terminal behaviors are analyzed as well. Stability criteria for several types of grid enabled by the SST are derived. The bidirectional power flow effect is also considered. These instability phenomena are demonstrated in ac, dc, and hybrid ac/dc grids of FREEDM system using HIL test bed. Finally, the conclusions are given and the scope of future work is discussed. / A Dissertation submitted to the Department of Electrical and Computer Engineering in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Fall Semester 2017. / September 7, 2017. / FREEDM system, harmonics, instability mitigation, paralleled inverters, power converter interraction, stability criterion / Includes bibliographical references. / Hui Li, Professor Directing Dissertation; Emmanuel G. Collins, University Representative; Mischa Steurer, Committee Member; Ming Yu, Committee Member.

High Temperature Superconductors (HTS) have the advantage of carrying direct current at zero resistance when operated below their critical temperature. At lower temperatures, these superconductors have the capability of carrying higher current densities. HTS power systems have applications in electrical power grids, defense, naval, aircraft, and industrial sectors. HTS devices enable higher efficiency while providing resiliency and reliability to power systems. This study developed models for superconducting cable system with two terminations, HTS cable, and cryo-cooler. The models combined electrical and cryogenic thermal aspects of the superconducting cable system. Several operating scenarios were simulated. Some contingencies such as cryo-cooler failure, circulation system failure were also modeled. A comparison of AC and DC cables was also analyzed in the system. The simulation models help in the analysis of the effects of system failure and to estimate the time required to turn off the system before the cable is affected. The results indicate that most of the heat load into the system is due to the terminations which are the interfaces between the superconducting cable and the room temperature components. In the contingency situations such as cryo-cooler failure, the time required to turn-off the system is several minutes. These results help us protect the cable from catastrophic damage during unexpected situations. Through these models, it is possible to calculate the maximum current that can be run through the system before the cable reaches a potential quench. / A Thesis submitted to the Department of Electrical and Computer Engineering in partial fulfillment of the requirements for the degree of Master of Science. / Fall Semester 2017. / November 14, 2017. / Includes bibliographical references. / Sastry V. Pamidi, Professor Directing Thesis; Simon Foo, Committee Member; Pedro Moss, Committee Member.

The Heisenberg Uncertainty principle is a fundamental concept from Quantum Mechanics that also describes the Fourier Transform. Unfortunately, it does not directly apply to the digital signals. However, it can be generalized if we use entropy rather than energy to form an uncertainty relation. This form of uncertainty, called the Hirschman Uncertainty, uses the Shannon Entropy. The Hirschman Uncertainty is defined as the average of the Shannon entropies of the discrete-time signal and its Fourier Transform. The functions that minimize this uncertainty are not the wellknown Gaussians from the Heisenberg theory, but are the picket fence functions first noticed in wavelet denoising. This connection suggests that the Hirschman Uncertainty is fundamental, but not conclusively. Here in this research, we develop two new uncertainty measures that are derived from the Hirschman Uncertainty. We want to use these measures to explore the fundamental nature of the Hirschman Uncertainty. In the first case, we replace the Shannon entropy with the Rényi entropy and study the impact of varying the Rényi order on the uncertainty of various digital signals. We call this new uncertainty measure, the Hirschman-Rényi uncertainty denoted by U[alpha over ½](x). We find that the derived uncertainty measure is invariant to the Rényi order in case of the picket fence signals and varies in case of other the digital signals like rectagular, cosine, square wave signals to name a few. This new uncertainty measure that utilizes the Rényi entropy decays with the increase in Rényi order value. Considering the invariance in uncertainty in case of picket fence signal, we can use either Shannon or Rényi entropy with any value of Rényi order to calculate Hirschman Uncertainty. In the second case, we derive an uncertainty measure that replaces the Fourier Transform with the Fractional Fourier Transform. The Hirschman Uncertainty using dFRT denoted by U[alpha over ½](x) is explored with the help of the minimizers of the Hirschman Uncertainty (the picket fence signals) along with the other digital signals. In this case, we find that the degree of rotation in the Fractional Fourier Transform does impact the uncertainty at the integer values of the transfer order. But for the non-interger values of the transfer order, the uncertainty variations are greatly reduced or are minimal. Finally to help verify our theory, we perform a classical texture recognition experiment. We find that the recognition performance follows directly as our developed Hirschman Rényi Uncertainty and the Hirschman Uncertainty using dFRT theory suggests. Additionally, it appears that a predictive solution for the proper selection of the Rényi order and the rotation angle can be developed that could significantly aid in image analysis. Our recognition results are consistent with entropic invariance theory in case of the two uncertainty measures. These results suggests that the Hirschman Uncertainty may be a fundamental characteristic of the digital signals. / A Dissertation submitted to the Department of Electrical and Computer Engineering in partial fulfillment of the requirements for the degree of Doctor of Philosophy. / Fall Semester 2015. / November 3, 2015. / Discrete fractional fourier transform, entropy, Hirschman Uncertainty, Texture classification, Uncertainty / Includes bibliographical references. / Victor DeBrunner, Professor Directing Dissertation; Anuj Srivastava, University Representative; Linda DeBrunner, Committee Member; Bruce Harvey, Committee Member; Rodney Roberts, Committee Member.