Engineering Commons^™

A Practical Continually Online Trained Artificial Neural Network Controller For A Turbogenerator, Ganesh K. Venayagamoorthy, Ronald G. Harley

Electrical and Computer Engineering Faculty Research & Creative Works

This paper reports on the simulation and practical studies carried out on a single turbogenerator connected to an infinite bus through a short transmission line, with a continually online trained (COT) artificial neural network (ANN) controller to identify the turbogenerator, and another COT ANN to control the turbogenerator. This identifier/controller augments/replaces the automatic voltage regulator and the turbine governor. Results are presented to show that this COT ANN identifier/controller has the potential to allow turbogenerators to operate more closely to their steady-state stability limits and nevertheless “ride through” severe transient disturbances such as three phase faults. This allows greater usage …

Multiple Model Adaptive Estimation And Control Redistribution Performance On The Vista F-16 During Partial Actuator Impairments, Curtis S. Clark

Theses and Dissertations

Multiple Model Adaptive Estimation with Control Reconfiguration (MMAE/CR) capability to estimate and compensate for partial actuator failures, or "impairments" is investigated using the high-fidelity, nonlinear, six-degree-of-freedom, VISTA F-16 simulation which currently resides on the Simulation Rapid-Prototyping Facility (SRF). After developing a model for inserting partial actuator impairments into the VISTA F-16 truth model, research begins with a battery of single actuator impairment tests. This stage of research explores the capability of the existing MMAE algorithm to estimate single, partial actuator impairments, and helps to define refinements and expansions needed in the MMAE algorithm for the second phase of research: the …

Comparison Of Fdtd Algorithms For Subcellular Modeling Of Slots In Shielding Enclosures, Kuang-Ping Ma, Min Li, James L. Drewniak, Todd H. Hubing, Thomas Van Doren

Electrical and Computer Engineering Faculty Research & Creative Works

Subcellular modeling of thin slots in the finite-difference time-domain (FDTD) method is investigated. Two subcellular algorithms for modeling thin slots with the FDTD method are compared for application to shielding end osures in electromagnetic compatibility (EMC). The stability of the algorithms is investigated, and comparisons between the two methods for slots in planes, and slots in loaded cavities are made. Results for scattering from a finite-length slot in an infinite plane employing one of the algorithms are shown to agree well with published experimental results, and power delivered to an enclosure with a slot agree well with results measured for …

Stability Of Discrete-Time Matrix Polynomials, Khanh Tu Ngo, Kelvin T. Erickson

Electrical and Computer Engineering Faculty Research & Creative Works

This paper derives conditions for the stability of discrete-time systems that can be modeled by a vector difference equation, where the variables are m x 1 vectors and the coefficients are m x m matrices. Stability of the system is related to the locations of the roots of the determinant of a real m x m matrix polynomial of nth order. In this case, sufficient conditions for the system to be stable are derived. The conditions are imposed on the infinity norm of two matrices constructed from the coefficient matrices and do not require the computation of the determinant polynomial. …

The Multirate Method For Simulation Of Power System Dynamics, J. G. Chen, Mariesa Crow

Electrical and Computer Engineering Faculty Research & Creative Works

In this paper, the multirate method is introduced to analyse power system behavior for systems with widely varying time constants. Estimates for computational speed-up are derived based on a linear system approach. The method is applied to a small power system example and the results, both in terms of accuracy and computation time, are compared to traditional simulation methods

Reduction Of Acoustic Feedback Oscillations By Use Of Spectrum Shifting, Youssef Abdelilah

Theses

The problem of whistling noise or self-oscillation in public address systems operating in closed halls or rooms can be very disturbing to the listeners. This noise is sustained when the positive acoustic feedback, of the public address-room system, meets the oscillation criteria. This feedback system results from the acoustic sound signal reflected off the room walls and any other obstacles in the room, originating from the public address speakers and then re-entering the microphone.

When all of the audio signal components entering the public address system are shifted by a frequency increment Δf of 6 Hz, an increase of 5 …

Coperite-Computer-Aided Tool For Power Engineering Research, Instruction, Training And Education, Badrul H. Chowdhury, D. E. Clark

Electrical and Computer Engineering Faculty Research & Creative Works

A graphics-oriented, primarily PC-based tool for education, research, and training in power engineering is introduced. The tool, called COPERITE, has all user interfaces resident on an IBM-386 microcomputer. Menus and windows are used for the interface, and attractive graphical representations and displays are included. Application programs that are interfaced are power flow, contingency analysis, economic dispatch, security-constrained dispatch, system stability, and fault analysis. These programs are executed on a VAX 8800 computer mainly for speed of execution. Information exchange between the PC and the VAX is made through an Ethernet connection which is transparent to the user. Results of execution …

The Parallel Implementation Of The Waveform Relaxation Method For Transient Stability Simulations, Marija D. Ilić, Mariesa Crow

Electrical and Computer Engineering Faculty Research & Creative Works

In this paper, the authors extend the results of their earlier paper on waveform relamtion (WR), which is a parallel algorithm for transient stability analysis. The WR algorithm is extended to a structure-preserving power system model in which the loads are retained. This results in a system of differential/ algebraic equations (DAEs). Power systems exhibit several unique dynamic properties which may be exploited in an advantageous manner by the WR algorithm. This leads to a greater computational efficiency than most other direct methods of simulation. This paper presents several theoretical results as well as computational results on parallel implementation.