Computational Methods for Electric Power Systems, Second Edition

Improve Compensation Strategies for Package Shortcomings In today's deregulated environment, the nation's electric power network is forced to operate in a manner for which it was not designed. As a result, precision system analysis is essential to predict and continually update network operating status, estimate current power flows and bus voltages, determine stability limits, and minimize costs. Computational Methods for Electric Power Systems is an introductory overview of computational methods used for analytical studies in power systems and other engineering and scientific fields. As power systems increasingly operate under stressed conditions, techniques such as computer simulation remain integral to control and security assessment. This volume analyzes the algorithms used in commercial analysis packages and presents salient examples of their implementation that are simple and thorough enough to be reproduced easily. Most of the examples were produced using MATLAB® language. Presents General Theory Applicable to Different Systems Commercial packages routinely fail or give erroneous results when used to simulate stressed systems, and understanding their underlying numerical algorithms is imperative to correctly interpret their results. This edition paints a broad picture of the methods used in such packages but omits extraneous detail. It includes new chapters that address function approximation and finite element analysis, in addition to new sections on: Generalized Minimal Residual (GMRES) methodsNumerical differentiationSecant methodHomotopy and continuation methodsPower method for computing dominant eigenvaluesSingular-value decomposition and pseudoinversesMatrix pencil method This book will enable users to make better choices and improve their grasp of the situations in which methods may fail—instilling greater confidence in the use of commercial packages.Table of ContentIntroduction The Solution of Linear Systems Gaussian Elimination LU Factorization Condition Numbers and Error Propagation Relaxation Methods Conjugate Gradient Methods Generalized Minimal Residual Algorithm (GMRES) Systems of Nonlinear Equations Fixed Point Iteration Newton-Raphson Iteration Continuation Methods Secant Method Numerical Differentiation Power System Applications Sparse Matrix Solution Techniques Storage Methods Sparse Matrix Representation Ordering Schemes Power System Applications Numerical Integration One-Step Methods Multistep Methods Accuracy and Error Analysis Numerical Stability Analysis Stiff Systems Step-Size Selection Differential-Algebraic Equations Power System Applications Optimization Least Squares State Estimation Linear Programming Nonlinear Programming Power System Applications Eigenvalue Problems The Power Method The QR Algorithm Arnoldi Methods Singular Value Decomposition Modal Identification Power System ApplicationsReview Quote"This is good resource material for a graduate student preparing for a qualifying exam"-- IEEE Power & Energy Magazine Biographical NoteDr. Mariesa Crow is a professor of electrical engineering at Missouri University of Science and Technology in Rolla. She is director of the Energy Research and Development Center. Her areas of research include: computer-aided analysis of power systems dynamics and security analysis, voltage stability, computational algorithms for analyzing stressed, non-linear, non-continuous systems. Power-electronic applications in bulk power systems (FACTS), and parameter estimation.