Engineering Commons^™

Transient Stability Estimation Using Artificial Intelligence Methodology: An Inductive Inference Approach, Almoataz Youssef Abdelaziz

Almoataz Youssef Abdelaziz

Many techniques are used for Transient Stability (TS) estimation of synchronous generators encompassing traditional time domain state numerical integration, Lyapunov based methods, probabilistic approaches and Artificial Intelligence (AI) techniques like pattern recognition and recently neural networks. This paper examines another recently proposed artificial intelligence technique to tackle the transient stability problem. The technique is based on the Inductive Inference Reasoning approach which belongs to a particular family of Machine Learning from Examples. Not only steady state but transient attributes are also used for transient stability estimation so as to reflect machine dynamics and network changes due to faults. The approach …

A Fuzzy Simulation Model Of Basic Commutating Electrical Machines, D. K. Chaturvedi, P. S. Satsangi, P. K. Kalra

D. K. Chaturvedi Dr.

Fuzzy Logic as applied to a great extent in controlling the process, plants and various complex systems, due to its inherent advantages like simplicity, ease in design, robustness and adaptivity. Aslo it is established that this approach works very well especially when the systems are not transparent. In this paper the approach is used for the modelling and simulation of an electrical machine to predict the behaviour of the machine under running conditionsas well as unde starting conditions. The starting characteristics of electrical machines are non-linear in nature, and it is very difficult to model them accurately. Also the developed …

Fitness Landscape Analysis Of Discrete Constrained Optimization Problems, Sai Prasanna Ravichandran

Masters Theses

The exponentially large size of the solution space of typical optimization problems precludes the use of any deterministic approach to search for an optimal solution. A more efficient and efficacious approach is to use heuristic algorithms based on rules of thumb to guide the search process in the solution space. Associated with every solution in the space is a real number called fitness that signifies the quality of a solution. This space and the fitness values together form the fitness landscapes. Knowledge about the topology of these fitness landscapes is vital for any heuristic search operator to expedite the search …

Adaptive Channel Equalization Techniques, Saied Razavilar

Masters Theses

Inter Symbol Interference (ISI) is a characteristic of band-limited communication channels that can severely degrade the digital communication system performance if not treated properly. Channel equalization techniques may be used to mitigate the effects of ISI in the system, thereby improving the system capacity and performance. In this thesis, the problem of Inter Symbol Interference caused by the nonlinear characteristics of band limited communication channels and adaptive equalization techniques to combat the effects of ISI are discussed. First, the analysis of ISI in a band limited communication channel using a concrete theoretical framework is presented. Next, various adaptive equalization techniques …

Multi-Baseline Interferometric Sar For Iterative Height Estimation, Adam E. Robertson

Theses and Dissertations

Multiple SAR interferograms with judiciously selected height sensitivities can be iteratively combined to create a high accuracy digital elevation map. An initial height estimate is refined by iteratively using larger baselines to obtain a height estimation accuracy limited by the spatial decorrelation of the antenna baseline. Spatial filtering is used to reduce the propagation of errors for accurate height estimation. Images containing regions isolated by phase discontinuities, as often found in urban environments, can be resolved by this iterative multi-baseline technique. Computationally demanding and potentially unreliable phase unwrapping is not required to determine scene elevation using SAR inMultiple SAR interferograms …

An Image Model Based On Occluding Object Images And Maximum Entropy, John A. Stuller, Rahul Shah

Electrical and Computer Engineering Faculty Research & Creative Works

This Paper Introduces a Statistical Image Model based on Occlusion and Maximum Entropy. the Statistical Model Combines a Fundamental Property of Image Formation, Occlusion, with Both Object-Image Shape and Nonuniform Object-Image Intensity. the Model is a Composition of Individual Object-Images that Have Random Positions, Shapes, and Intensities, and that Occlude Both Background and One Another. We Derive the Autocorrelation and Second-Order Probability Density Functions of This Model and Give Several Examples. © 1998 IEEE.

What Can Humor And Advertising Teach Us About Legal Drafting?, Dallin D. Oaks, Jeremy Lewis

Faculty Publications

In his book Legal Drafting in a Nutshell, Haggard includes a useful chapter of nearly 50 pages in which he lists various lexical and structural elements and explains how these can lead to ambiguous writing. The kind of inventory approach that he presents is helpful and reveals that ambiguity doesn't just happen but often results from the careless use of identifiable structures or particular words.

A Recursive Green's Function Method For Boundary Integral Analysis Of Inhomogeneous Domains, Jim D. Freeze, Michael A. Jensen

Faculty Publications

The recursive Green's function method (RGFM) for computation of fields scattered by two-dimensional (2-D) inhomogeneous dielectric bodies is presented. The algorithm efficiently constructs the Green's function for the inhomogeneous region by recursively combining known Green's functions from smaller subdomains. The fields on the scatterer surface are then computed using a boundary integral formulation. Proper implementation of the RGFM results in computational and storage complexities which scale as N1.5 and N, respectively, where N is the total number of discrete cells in a domain. Comparisons of results obtained using the RGFM with those computed from moment method and exact solutions show …

Primary Research Of Reversible Proton Exchange Membrane Regenerative Fuel Cells, Zhigang Shao, Baolian Yi, Ming Han

Journal of Electrochemistry

In this paper, we report the results of our studies on regenerative proton exchange membrane fuel cell(RPEMRFC) catalysts, electrode preparation and their performances. The Pt catalyst for the oxygen electrode was replaced by 60 w%Pt+40 w%IrO 2 which reduces polarization by 0.2 V in a polymer protone exchange membrane fuel cell. The cycle performance of RPEMRFC was found to be manly effected by the stability of the diffiusion layer of bifunctional oxygen electrode.

The Latest World Wide Progress In Solid Oxide Fuel Cell, Jiang Yi

Journal of Electrochemistry

Solid oxide fuel cell(SOFC) is a novel, green electricity generation technology, which has been showing rapid development since 1980's. In light of its high efficiency, environmental friendly, fuel adaptability and long term stability, the Research, Development and Demonstration(RD&D) of this hype of full cell have been emphasized in a large number of developed countries. Based on the information communicated at the latest symposium on SOFC held last year in Aachen(Germany) and the knowledge gained through years of research in this field, a comprehensive review on the latest development of the tubular, this planar SOFC as well as on a number …

Study On Electrosynthesis Of Butyl Butyrate At Conducting Oxide Anode Support On Titanium, Junxiang He, Jincheng Zhou

Journal of Electrochemistry

Synthesis of butyl butyrate by direct electroxidation of n butyl alcohol and the electrochemical behaviors of several selected conducting oxide electrodes (Pb/PbO 2、Ti/PbO 2、i/SnO 2 +Sb 2O 3 /PbO 2 and Ti/SnO 2+Sb 2O 3+PdO x/PbO 2) have been investigated. The results demonstrated that the intermediate layers of SnO 2、Sb 2O 3 and PdO x could significantly decrease the interfacial resistance of the electrodes,and that the Ti/SnO 2+Sb 2O 3+PdO x/PbO 2 electrode possessed better electrocatelytic performance than others.

Studies Of Spinal Li 1+X Mn 2-X O 4 Rechargeable Lithium Ion Battery, Xiaomei Wu, Qinghe Yang, Zhongkao Jin, Haoqing Wu, Xiangfu Zong

Journal of Electrochemistry

The spinel Li intercalation compound Li 1+ x Mn 2- x O 4(0< x <1) was prepared by using techniques of cyclic voltammetry. The electrochemical parameters of Li 1+ x Mn 2- x O 4 electrode vary as either the starting materials or the synthesis temperature changes. It was found that the initial capacity and reversible behaviors of the Li 1+ x Mn 2- x O 4 electrode depends on the amount of the excess Li, represented as x in Li 1+ x Mn 2- x O 4. The reversible of Li intercalation process for Li 1+ x Mn 2- x O 4 electrode was demonstrate by cyclic voltammograms.

Drect Electro_Oxidation Synthesis Of D Arabinose On Spectroscopic Graphite Electrode, Hei Wei, Huang Ying

Journal of Electrochemistry

D Arabinose synthesized by direct electro_oxidative decarboxylation of D gluconate on graphite electrode from aqueous solutions. The effects of D gluconate concentration, electrolysis potential, temperature and electrolyte composition on the current efficiency and product yield of electrosynthsis are discussed. The laboratory scale electrosynthesis was achieved with flow through cell.The optimum technology conditions for electrosynthesis were obtained with orthogonal design method.Both galvanostatic and potentionstatic conditions have been exploited, the later was found to be more effective.

A Preliminary Study Of The Synthesis Of Tin Based Amorphous Material And Its Behavior On Lithium Intercalation, Li Liu, Hanxi Yang, Jutang Sun, Xingping Ai

Journal of Electrochemistry

An amorphous lithium inserting anode material based on Tin oxide glass was synthesized by salification in solution and its properties on lithium intercalation was investigated by X ray diffraction(XRD), microelectrode voltammetry and charge discharge measurements. The results have shown that this kind of material has considerable electrochemical reversibility toward lithium intercalation deintercalation. The available discharge capacity of the material reaches 430 mA·h/g with charge_discharge efficiency over 90%, showing a great potential for practical use.

Self-Enhancement Of Dynamic Gratings In Photogalvanic Crystals, Nickolai Kukhtarev, Sergei F. Lyuksyutov, Preben Buchhave, Tatiana Kukhtareva, K. Sayano, Partha P. Banerjee

Electrical and Computer Engineering Faculty Publications

We have developed a compact closed-form solution of the band transport model for high-contrast gratings in photogalvanic crystals. Our solution predicts the effect of the photoconductivity and the electric field grating enhancement due to the photogalvanic effect. We predict a pronounced dependence of the steady-state photogalvanic current on the contrast of the interference pattern and an increase of holographic storage time due to the enhancement of the photoconductivity grating contrast. In the high contrast limit and a large photogalvanic effect the refractive index grating will be shifted from the position of the intensity modulation pattern, contrary to the usually adopted …

A Virtual Classroom Via Dataconferencing: A Multi-Institutional Experience, Shelli K. Starrett, Anil Pahwa, Kraig J. Olejniczak, S. D. Sudhoff, Mariesa Crow

Electrical and Computer Engineering Faculty Research & Creative Works

This paper presents the experience of five faculty from four universities who are co-developing and team teaching courses using the world-wide web (WWW). In particular two novel power engineering courses, Flexible Control of Transmission Systems and Flexible Control of Distribuction Systems, introduce the student to the application of power electronic-based controllers, known specifically as flexible AC transmission systems (FACTS) controllers, to enhance the performance of the transmission and distribution systems in an electric power system. In order to leverage faculty expertise in the various topics included in each course, a unique partnership was formed among five faculty from the following …

An Intelligent Reactive Controller For Ocean Science Autonomous Underwater Vehicles, Roy M. Turner

University of Maine Office of Research Administration: Grant Reports

Most of the research activities performed during this grant period had to do with designing and implementing the Orca intelligent autonomous underwater vehicle (AUV) controller. This entailed a requirements analysis early in the project, followed by the design and implementation of the program. The latter two activities were interleaved; Orca was designed and implemented in a phased manner, with each new version of the program having more functionality.

Deriving Of Single Intensive Picosecond Optical Pulses From A High-Power Gain-Switched Laser Diode By Spectral Filtering, S. N. Vainshtein, Grigory Simin, J. T. Kostamovaara

Faculty Publications

Single 25 ps/16 W optical pulses were achieved by spectral filtering from a multiheterostructure gain-switched laser diode with its quasisteady-state modes suppressed by a factor of 10³ as compared with the peak power. A significant transient spectrum broadening makes this possible provided that a very high dI/dt rate of the pumping current pulse is used. A simple numerical model is suggested which describes adequately both the spectral and transient features of the observed phenomenon. It follows from the model that single picosecond optical pulses can be obtained from any type of high power semiconductor laser.

Extension Of The Generalized Hebbian Algorithm For Principal Component Extraction, Fredric M. Ham, Inho Kim

Electrical Engineering and Computer Science Faculty Publications

Principal component analysis (PCA) plays an important role in various areas. In many applications it is necessary to adaptively compute the principal components of the input data. Over the past several years, there have been numerous neural network approaches to adaptively extract principal components for PCA. One of he most popular learning rules for training a single-layer linear network for principal component extraction is Sanger's generalized Hebbian algorithm (GHA). We have extended the GHA (EGHA) by including a positive-definite symmetric weighting matrix in the representation error-cost function that is used to derive the learning rule to train the network. The …

Validation Of Waimss Incident Duration Estimation Model, Wei Wu, Pushkin Kachroo, Kaan Ozbay

Electrical & Computer Engineering Faculty Research

This paper presents an effort to validate the traffic incident duration estimation model of WAIMSS (wide area incident management support system). Duration estimation model of WAIMSS predicts the incident duration based on an estimation tree which was calibrated using incident data collected in Northern Virginia. Due to the limited sample size, a full scale test of the distribution, mean and variance of incident duration was performed only for the root node of the estimation tree, white only mean tests were executed at all other nodes whenever a data subset was available. Further studies were also conducted on the model error …

Wide-Area Incident Management System On The Internet, Kaan Ozbay, Pushkin Kachroo

Electrical & Computer Engineering Faculty Research

The incident management process consists of four sequential steps-incident detection, response, clearance and recovery. Each of these components comprises of a number of operations and coordinated decision-making between the agencies involved. The provision of computer based support tools for the personnel involved will help develop appropriate strategies and increase efficiency and expediency. Existing systems are developed on various traditional computing platforms. However, with the advent of World Wide Web and Internet based programming tools such as Java, it is now possible to develop platform independent decision support tools for the incident management agencies. Any agency will be able to use …

Modeling And Simulation Of High-Field, High-Temperature Sic Devices, Sanjay Pathak

Electrical & Computer Engineering Theses & Dissertations

With superior properties such as a large band gap, high thermal conductivity, and large electron drift velocity, SiC is expected to give a new dimension to high power, high temperature electronic devices used in power applications, microwave circuits, and for the space and automobile industries. SiC possesses an inherent advantage over its other large bandgap competitors in terms of inherited processing and device technology from Si. However, there is inadequate understanding of SiC device and its parameters. Simple extrapolations from Si are known to be inadequate. The aim of this dissertation, therefore. is to produce better understanding of SiC devices …

Incorporating Vertical Discontinuities In Power-Bus Modeling Using A Mixed-Potential Integral Equation And Circuit Extraction Formulation, Jun Fan, Hao Shi, James L. Drewniak, Todd H. Hubing, Richard E. Dubroff, Thomas Van Doren

Electrical and Computer Engineering Faculty Research & Creative Works

Noise on the DC power-bus attributed to device switching is among the primary sources of EMI and signal integrity problems. A mixed-potential integral equation formulation with circuit extraction approach is used to model the planar multi-layer power-bus geometry, which can also include arbitrary shaped power regions on multiple layers. Incorporating vertical discontinuities, e.g., decoupling capacitor interconnects, is a critical aspect of the modeling, and must be done properly since they are included as a lumped element model and not a part of the MPIE formulation. Agreement with experimental results demonstrate the present approach.

New Cmos Universal Constant-Gm Input Stage, Vladimir I. Prodanov, Michael M. Green

Electrical Engineering

Conceptually new constant-Gm input-stage architecture is presented. It provides near constant net transconductance independent of transistor operating region-strong, moderate or weak inversion. One possible implementation of the proposed architecture is discussed in details. Its operation has been experimentally verified. Results from performed measurements are included.

A Neural Network Pruning Algorithm With Embedded Gradient-Conjugate Training For The Identification Of Large Flexible Space Structures, Xiao-Hua Yu

Electrical Engineering

The choice of network dimension is a fundamental issue in the design of artificial neural networks. A larger neural network is powerful for solving problems while a smaller neural network is always advantageous in real-time environment where speed is crucial. In this paper, a network pruning algorithm with embedded gradient-conjugate training is investigated and applied to the identification of a large flexible space structure. Computer simulation results show that this approach can dramatically reduce the size of neural network while maintaining compatible identification accuracy.

Development Of The Industrial Expert System And Its Application For The Rational Choice Of Crushing And Grinding Facilities Of A Cement Factory (Développement Du Système Expert Industriel Et Son Application Pour Le Choix Rationnel Des Installations De Concassage Et De Broyage D'Une Cimenterie), Boukhemis Chetate , Cherif Kara, Galina Tcheremechko, Irina Merazka

Boukhemis Chetate

Les économies d’énergie électrique suscitent un intérêt croissant dans le domaine de l’industrie des matériaux de construction. En effet, la dépense en électricité peut représenter jusqu’à 50% de l’ensemble des coûts d’énergie nécessaires à la fabrication d’une tonne de ciment. Par ailleurs, environ 80% de l’énergie utilisée pour la fabrication d’une tonne de ciment est dépensée pour la fragmentation des matériaux de construction, c'est-à-dire pour le concassage et le broyage. Par conséquent, le choix rationnel des installations de concassage et de broyage et l’optimisation de leurs régimes de fonctionnement représentent un facteur important d’économies d’énergie. Toutefois, il est de plus …

Production And Propagation Of Hermite–Sinusoidal-Gaussian Laser Beams, Lee W. Casperson, Anthony A. Tovar

Electrical and Computer Engineering Faculty Publications and Presentations

Hermite–sinusoidal-Gaussian solutions to the wave equation have recently been obtained. In the limit of large Hermite–Gaussian beam size, the sinusoidal factors are dominant and reduce to the conventional modes of a rectangular waveguide. In the opposite limit the beams reduce to the familiar Hermite–Gaussian form. The propagation of these beams is examined in detail, and resonators are designed that will produce them. As an example, a special resonator is designed to produce hyperbolic-sine-Gaussian beams. This ring resonator contains a hyperbolic-cosine-Gaussian apodized aperture. The beam mode has finite energy and is perturbation stable.

Automatic Target Cueing Of Hyperspectral Image Data, Terry A. Wilson

Theses and Dissertations

Modern imaging sensors produce vast amounts data, overwhelming human analysts. One such sensor is the Airborne Visible and Infrared Imaging Spectrometer (AVIRIS) hyperspectral sensor. The AVIRIS sensor simultaneously collects data in 224 spectral bands that range from 0.4µm to 2.5µm in approximately 10nm increments, producing 224 images, each representing a single spectral band. Autonomous systems are required that can fuse "important" spectral bands and then classify regions of interest if all of this data is to be exploited. This dissertation presents a comprehensive solution that consists of a new physiologically motivated fusion algorithm and a novel Bayes optimal self-architecting classifier …

An Efficient And Effective Approach To Column-Based Input/Output Encoding In Functional Decomposition, Michael Burns, Marek Perkowski, Stanislaw Grygiel, Lech Jozwiak

Electrical and Computer Engineering Faculty Publications and Presentations

Encoding in Curtis-style decompositions is the process of assigning codes to groups of compatible columns (or cubes) so that the binary logic descriptions of the predecessor and successor sub-functions can be created for further decomposition. In doing so, the sub-functions created are functionally equivalent to the set of care values specified in the original function. In this paper an input/output encoding algorithm DC_ENC is presented that is designed to achieve the simplest total complexity of the predecessor and successor sub-functions, and to increase the total number of don't cares for their further utilization in subsequent decomposition steps of these sub-functions.

Automatic Analysis Of Electronic Drawings Using Neural Network, Yi Shi

Dissertations

Neural network technique has been found to be a powerful tool in pattern recognition. It captures associations or discovers regularities with a set of patterns, where the types, number of variables or diversity of the data are very great, the relationships between variables are vaguely understood, or the relationships are difficult to describe adequately with conventional approaches.

In this dissertation, which is related to the research and the system design aiming at recognizing the digital gate symbols and characters in electronic drawings, we have proposed: (1) A modified Kohonen neural network with a shift-invariant capability in pattern recognition; (2) An …