<p> Wireless power transfer is an emerging technology for electric and plug-in hybrid electric vehicles, addressing the power needs of these vehicle's charging systems. The current project presents a detailed technical analysis of an existing wireless charging system, and proposes a new design for a significant improvement in performance. Simulation results and comparisons between the two systems show that the proposed design will shorten the time to fully recharge a hybrid vehicle battery by 30% to 50% as compared to the existing system. </p>

<p> This project report presents a Field Programmable Gate Array (FPGA) implementation of an adaptive receiver designed and patented by Dr. Rajendra Kumar, U.S. patent 8233568, July 31, 2012. This adaptive receiver is generally used in the systems where higher order modulated signals are used for transmitting information over a fading a channel. </p><p> In communication systems, a receiver design must be reliable. Previously, the adaptive receivers made use of pilot signals in order to correct the phase of the transmitted signal. However, this system faced challenges and failed to accurately detect the higher order modulated signals. To eliminate this challenge, the adaptive receiver designed and patented by Dr. Rajendra Kumar estimates the channel fade and the phase using an estimator, phase detector, and Kalman filter that decodes the data and provides the required channel equalization without the need of any pilot symbols. </p><p> This project efficiently implements Rajendra Kumar&rsquo;s Adaptive Receiver on FPGA with reduced number of gates. Very High Speed Integrated Circuit Hardware Description Language (VHDL) codes and Xilinx ISE are used for replicating the adaptive receiver&rsquo;s circuit on Nexys 3 Spartan-6 FPGA trainer board. </p>

<p> Virtual reality (VR) is a computer generated simulation of a 3D environment. It artificially creates a sensory experience that allows the user to see the virtual environment. Live video streaming &ldquo;virtual reality&rdquo; is a new technology that would transport a scene from the real environment to people who are unable to visit in person, and experience it up close and in real-time. The experience is interactive, as the viewers have the freedom to look around and move the camera position and angle to see the real environment in any direction during the live-broadcast. In this project, a single board computer is used to capture and stream video. A custom video player converts this 2D video to virtual 3D video for VR headsets. Users can now use a smartphone with a VR headset to view, control, and enjoy live video streaming from anywhere in the world with an internet connection.</p>

<p> Pedestrian detection and counting have important application in video surveillance for entrance monitoring, customer behavior analysis, and public service management. In this thesis, we propose an accurate, reliable and fast method for pedestrian detection and counting in video surveillance. To this end, we first develop an effective method for background modeling, subtraction, update, and shadow removal. To effectively differentiate person image patches from other background patches, we develop a head-shoulder classification and detection method. A foreground mask curve analysis method is to determine the possible position of persons, and then use a SVM (Support Vector Machine) classifier with HOG (Histogram of Oriented) feature and bag of words to detect the head-shoulder of people. Based on the foreground detection and head-shoulder classification at each frame, we develop a person counting algorithm in the temporal domain to analyze the frame-level classification results. Our experiments with real-world surveillance videos demonstrate the proposed method has achieved accurate and reliable pedestrian detection and counting.</p>

<p> Rapid developments in the communications field have created a rising demand for low power, high speed, and low weight communication devices. The current project presents the development of a Viterbi decoder on a chip with a reduced dynamic power consumption, achieved by using an asynchronous design, which is data driven and active only when needed. The Xpower analyzer tool is used to measure the dynamic power on the designed chip, from which it is seen that the proposed design greatly improves power consumption. The results also show that there is a trade-off between dynamic power reduction, larger chip area, and reduced speed. </p>

<p> This research introduces two new methodologies to design a set of controllers such that every controller in the set preserves closed-loop stability of a given multi variable plant under prescribed loop failures. The proposed approaches differ from existing techniques in two ways: First, these methods are strictly based on frequency response data of the plant that can be easily measured by experiments. No mathematical models or system identification processes are used. Second, while most control design methods find one controller, the proposed methods design a set of controllers satisfying the control objective. Two approaches are presented with examples illustrating the controller design. Integrity test results of the designed controllers under pre-specified loop failures are also presented.</p>

<p>Current state of the art techniques for landmine detection in ground penetrating radar (GPR) utilize statistical methods to identify characteristics of a landmine response. This research makes use of 2-D slices of data in which subsurface landmine responses have hyperbolic shapes. Various methods from the field of visual image processing are adapted to the 2-D GPR data, producing superior landmine detection results. This research goes on to develop a physics-based GPR augmentation method motivated by current advances in visual object detection. This GPR specific augmentation is used to mitigate issues caused by insufficient training sets. This work shows that augmentation improves detection performance under training conditions that are normally very difficult. Finally, this work introduces the use of convolutional neural networks as a method to learn feature extraction parameters. These learned convolutional features outperform hand-designed features in GPR detection tasks. This work presents a number of methods, both borrowed from and motivated by the substantial work in visual image processing. The methods developed and presented in this work show an improvement in overall detection performance and introduce a method to improve the robustness of statistical classification.</p> / Dissertation

<p>While molecular and cellular processes are often modeled as stochastic processes, such as Brownian motion, chemical reaction networks and gene regulatory networks, there are few attempts to program a molecular-scale process to physically implement stochastic processes. DNA has been used as a substrate for programming molecular interactions, but its applications are restricted to deterministic functions and unfavorable properties such as slow processing, thermal annealing, aqueous solvents and difficult readout limit them to proof-of-concept purposes. To date, whether there exists a molecular process that can be programmed to implement stochastic processes for practical applications remains unknown. </p><p>In this dissertation, a fully specified Resonance Energy Transfer (RET) network between chromophores is accurately fabricated via DNA self-assembly, and the exciton dynamics in the RET network physically implement a stochastic process, specifically a continuous-time Markov chain (CTMC), which has a direct mapping to the physical geometry of the chromophore network. Excited by a light source, a RET network generates random samples in the temporal domain in the form of fluorescence photons which can be detected by a photon detector. The intrinsic sampling distribution of a RET network is derived as a phase-type distribution configured by its CTMC model. The conclusion is that the exciton dynamics in a RET network implement a general and important class of stochastic processes that can be directly and accurately programmed and used for practical applications of photonics and optoelectronics. Different approaches to using RET networks exist with vast potential applications. As an entropy source that can directly generate samples from virtually arbitrary distributions, RET networks can benefit applications that rely on generating random samples such as 1) fluorescent taggants and 2) stochastic computing.</p><p>By using RET networks between chromophores to implement fluorescent taggants with temporally coded signatures, the taggant design is not constrained by resolvable dyes and has a significantly larger coding capacity than spectrally or lifetime coded fluorescent taggants. Meanwhile, the taggant detection process becomes highly efficient, and the Maximum Likelihood Estimation (MLE) based taggant identification guarantees high accuracy even with only a few hundred detected photons.</p><p>Meanwhile, RET-based sampling units (RSU) can be constructed to accelerate probabilistic algorithms for wide applications in machine learning and data analytics. Because probabilistic algorithms often rely on iteratively sampling from parameterized distributions, they can be inefficient in practice on the deterministic hardware traditional computers use, especially for high-dimensional and complex problems. As an efficient universal sampling unit, the proposed RSU can be integrated into a processor / GPU as specialized functional units or organized as a discrete accelerator to bring substantial speedups and power savings.</p> / Dissertation

Cyber-physical systems are complex engineering systems that integrate computational, communication, and control components with physical components in many applications such as automotive systems, aeronautical systems, industrial process control systems, electrical power grids, and environmental monitoring systems. As the cyber components increase in both number and complexity, technical challenges arise for their integration with the physical domain. As the field of cyber-physical systems continues to grow and evolve, problems emerge from the interaction of heterogeneous domains, hybrid dynamics, and nonlinearities which significantly hamper the system integration. Consequently, rigorous engineering methods are needed for the integration of cyber and physical components in order to achieve predictable, correct behavior. This dissertation presents a model-based design framework based on port-Hamiltonian systems and passivity in order to address the challenges mentioned above. The contributions are threefold: (1) A domain-specific modeling language, (2) a compositional model-based control design method, and (2) a formal safety analysis method for multi-modal port Hamiltonian systems. The Port-Hamiltonian Systems Modeling Language uses the structure of port-Hamiltonian systems to model cyber-physical systems with nonlinearities, hybrid dynamics, and heterogeneous domains in a component-based way. The compositional model-based control design method uses passivity-based methods to ensure stability properties of the overall system in the presence of implementation uncertainties. The safety analysis method utilizes the Hamiltonian function as a barrier function to prevent system trajectories from ending in unsafe regions of the state space. The theoretical contributions are evaluated and validated with an in-depth case study of automotive control software for an autonomous vehicle using a hardware-in-the-loop simulation platform.

There has been a growing interest in electrochemical storage devices such as batteries, fuel cells and supercapacitors in recent years. This interest is due to our increasing dependence on portable electronic devices and on the high demand for energy storage from the electric transport vehicles and electrical power grid industries. As we transition towards cleaner renewable fuel sources such as solar, wind, tidal, etc. our dependence on energy storage devices will continue to grow. Li-air offers much higher energy density than all other batteries based on electrochemical storage. However, these batteries currently suffer from a number of issues such as a low cyclability and a reduced practical energy density compared to the theoretical energy density. The deposition of lithium peroxide on the surface of the cathode is one of the main causes for the low practical specific capacity of lithium-air batteries with organic electrolyte. Electrochemical impedance spectroscopy (EIS) has been used in the past to extract physical parameters such as chemical diffusion coefficient, effective diffusion coefficient, Faradaic reaction rate, degradation and stability of an electrochemical device. In this dissertation, a physics based analytical model is developed to study the EIS of Li-air batteries, in which the mass transport inside the cathode is limited by oxygen diffusion, during charge and discharge. The model takes into consideration the effects of double layer, Faradaic processes, and oxygen diffusion in the cathode, but neglects the effects of anode, separator, conductivity of the deposit layer, and Li-ion transport. The analytical model predicts that the effects of Faradaic impedance can be hidden by the double layer capacitance. Therefore, the dissertation focuses separately on two cases: 1) the case when the Faradaic process and the double layer capacitance are separate and can be observed as two different semicircles on the Nyquist plot and 2) the case when the Faradaic process is shadowed by the double layer capacitance and shows up as only one large semicircle on the Nyquist plot. A simple expression is developed to extract physical parameters such as the values of the diffusion coefficient of oxygen and Faradaic reaction rate from experimental impedance spectrum for each of the two cases. The diffusion coefficient can be determined by using the resistances (real impedance intercept on the Nyquist plot) of both the semicircles for the first case and by using the combined resistance for the second case. Once, the effective oxygen diffusion coefficient is estimated, it can be used to estimate the value of the reaction constant. This method of extracting the values of the diffusion coefficient and reaction constant can serve as a tool in identifying an effective electrolyte or cathode material. It can also serve as a noninvasive technique to identify and also quantify the use of the catalyst to improve the reaction kinetics in an electrochemical system. Finally, finite element simulations are used to validate the analytical models and to study the effects of discharge products on the impedance spectra of Li-air batteries with organic electrolyte. The finite element simulations are based on the theory of concentrated solutions and the complex impedance spectra are computed by linearizing the partial differential equations that describe the mass and charge transport in Li-air batteries. These equations include the oxygen diffusion equation, the Li drift-diffusion equation, and the electron conduction equation. The reaction at the anode and cathode are described by Butler-Volmer kinetics. The total impedance of a Li-air battery increases by more than 200% when the response is measured near the end of the discharge cycle as compared to on a fresh battery. The resistivity of the deposition layer significantly affects the deposition profile and the total impedance. Using electrolytes with high oxygen solubility and concentrated O2 gas at high pressures will reduce the total impedance of Li-air batteries. / 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. / August 26, 2015. / analytical modeling, diffusion impedance, EIS, impedance spectroscopy, Li-air / Includes bibliographical references. / Petru Andrei, Professor Directing Dissertation; Joseph B. Schlenoff, University Representative; Jim P. Zheng, Committee Member; Pedro Moss, Committee Member; Hui Li, Committee Member.