Surendra Singh, Klaus Halterman, and J. Merle Elson
Naval Air Warfare Center Weapons Division TP 8590
The bi-conjugate gradient (bi-CG) algorithm is applied to numerically solve linear equation systems resulting from integral equations arising in electromagnetic scattering problems. The basic advantage of using this algorithm over traditional methods, such as matrix inversion, is that the algorithm is iterative in nature . The iterative nature allows the user to control the residual error in the final solution. Also, the algorithm can be implemented without storing the coefficient matrix, thus providing huge saving in storage requirements. It was realized that the existing code that utilized matrix inversion to solve the linear equation system was limited to a coarse discretization of the geometry. This code could not handle very fine geometry discretization due to storage limitations. With the implementation of the bi-CG algorithm, this limitation was overcome . The present code can very easily handle very fine discretizations, thereby vastly improving the utility of the computer code. This report outlines the bi-CG algorithm and provides its implementation to solve an electromagnetic scattering problem of a nanowire illuminated by a plane wave. The report also includes a complete FORTRAN listing of the code.
Wilfred Kaplan and Donald J. Lewis
University of Michigan
The principal purpose of the book was to provide an integration of linear algebra and calculus. Chapters and topics are as follows:
- 0: Introduction, Review of Algebra, Geometry and Trigonometry
- 1: Two-Dimensional Vector Geometry
- 2: Limits
- 3: Differential Calculus
- 4: Integral Calculus
- 5: Elementary Transcendental Functions
- 6: Applications of Differential Calculus
- 7: Applications of Integral Calculus
- 8: Infinite Series
- 9: Vector Spaces
- 10: Matrices and Determinants
- 11: Linear Euclidean Geometry
- 12: Differential Calculus of Functions of Several Variables
- 13: Integral Calculus of Functions of Several Variables
- 14: Ordinary Differential Equations
Robert Michael Lewis, Virginia Torczon and Michael W. Trosset
ICASE Report No. 2000-26
We discuss direct search methods for unconstrained optimization. We give a modern perspective on this classical family of derivative-free algorithms, focusing on the development of direct search methods during their golden age from 1960 to 1971. We discuss how direct search methods are characterized by the absence of the construction of a model of the objective. We then consider a number of the classical direct search methods and discuss what research in the intervening years has uncovered about these algorithms. In particular, while the original direct search methods were consciously based on straightforward heuristics, more recent analysis has shown that in most—but not all—cases these heuristics actually suffice to ensure global convergence of at least one subsequence of the sequence of iterates to a first-order stationary point of the objective function.
18 January 2010
We model our world with continuous mathematics. Whether our interest is natural science, engineering,
even finance and economics, the models we most often employ are functions of real variables. The equations can be linear or nonlinear, involve derivatives, integrals, combinations of these and beyond. The tricks and techniques one learns in algebra and calculus for solving such systems exactly cannot tackle the complexities that arise in serious applications. Exact solution may require an intractable amount of work; worse, for many problems, it is impossible to write down an exact solution using elementary functions like polynomials, roots, trig functions, and logarithms.
This course tells a marvellous success story. Through the use of clever algorithms, careful analysis, and speedy computers, we are able to construct approximate solutions to these otherwise intractable problems with remarkable speed. Trefethen defines numerical analysis to be ‘the study of algorithms for the problems of continuous mathematics’. This course takes a tour through many such algorithms, sampling a variety of techniques suitable across many applications. We aim to assess alternative methods based on both accuracy and efficiency, to discern well-posed problems from ill-posed ones, and to see these methods in action through computer implementation.