Article

Article title MODIFICATION OF THE RANDOM SEARCH METHOD FOR STIFF OPTIMIZATION PROBLEMS
Authors V.N. Biryukov, P.N. Zubkov, A.M. Pilipenko
Section SECTION II. METHODS, MODELS AND ALGORITHMS OF INFORMATION PROCESSING
Month, Year 03, 2015 @en
Index UDC 519.85+621.382
DOI
Abstract The methods of solution of the important practical optimization problem – the object parameters identification from the measurements are considered. In this paper the object of research is the semiconductor diode which was intended for application in integrated balanced and bridge circuits. One of the most important requirements for the specified circuits is a high accuracy of the parameters determination for their elements. The objective of this work is to improve the accuracy and speed of parameters optimization for the studied object. The modification of the random search method, based on application of the nonuniform randomization, was proposed to ensure the work objective. The problems of parameters identification of simple two-parameter Shockley model and three-parameter SPICE-model of diode were solved for the proof the effectiveness of the proposed modified random search method. Application of Quasi-Newton methods and Levenberg-Marquardt method for solving of this problem does not allow to receive even a crude estimate of the parameters for the initial approximations that differ from the exact solution is no more than twice. It is possible to determine the diode parameters by random search method with reasonable accuracy but sufficiently large time of analysis. Application of the modified random search method with the nonuniform distribution law of random variables can increase the rate of descent and reduce the error of the results.

Download PDF

Keywords Nonlinear programming; optimization; random search; stiff systems; convergence analysis.
References 1. Kolda T.G., Lewis R.M., Torczon V. Optimization by Direct Search: New Perspectives on Some Classical and Modern Methods // SIAM Review. – 2003. –Vol. 45, N. 3. – P. 385-482.
2. Федоренко Р.П. Введение в вычислительную физику: Учеб. пособие для вузов. – М.: Изд-во МФТИ, 1994. – 528 с.
3. Альшина Е.А., Болтнев А.А., Качер О.А. Градиентные методы с ускоренной сходимостью // Журнал вычислительной математики и математической физики. – 2005. – Т. 45, № 3. – С. 374-382.
4. Черноруцкий И.Г. Методы оптимизации. Компьютерные технологии. – СПб.: БХВ-Петербург, 2011 – 384 с.
5. Черноруцкий И.Г. Практическая оптимизация и невыпуклые задачи // Научно-технические ведомости Санкт-Петербургского государственного политехнического университета. Информатика. Телекоммуникации. Управление. – 2013. – № 176. – С. 79-86.
6. Ракитский Ю. В., Устинов С. М., Черноруцкий И. Г. Численные методы решения жестких систем. – М.: Наука, 1979. – 208 с.
7. Tsividis Y.P., Suyama K. MOSFET modeling for analog circuit CAD: Problems and prospects // IEEE Journal of Solid-State Circuits. – 1994. – Vol. 29, N. 3. – P. 210-216.
8. Пилипенко А.М. Об обусловленности задач параметрической идентификации // Известия ТРТУ. – 2004. – № 8 (43). – С. 13-18.
9. Jaeger R.C., Blalock T.N. Microelectronic circuit design. Fourth Edition. – McGraw-Hill, New York, 2011. – 1365 p.
10. Fourer R., Gay D.M., Kernighan B.W. AMPL: A Modeling Language for Mathematical Programming, 2nd ed. - Thomson/Brooks/Cole, Pacific Grove, CA, 2003. – 517 p.
11. Kelley C.T. Iterative Methods for Optimization. – SIAM, Philadelphia, 1999. – 180 p.
12. Fermi E., Metropolis N. Numerical solution of a minimum problem. – Los Alamos, N.M.: Los Alamos Scientific Laboratory of the University of California, 1952. – 17 p.
13. Brent R.P. Algorithms for Minimization without Derivatives. – Dover Publications, Mineaola, NY, 2002. – 195 p
14. Карманов В.Г. Математическое программирование: Учеб. пособие. – 5-е изд., стереотип. – М.: Физматлит, 2004. – 264 с.
15. Растригин Л.А. Адаптация сложных систем: методы и приложения. – Рига: Зинатне, 1981. – 376 с.
16. Lewis R.M., Torczon V., Trosset M.W. Direct search methods: Then and now // Journal of Computational and Applied Mathematics. – 2000. – Vol. 124, N. 1-2. – P. 191-207.
17. Luus R., Jaakola T. H. I., Optimization by direct search and systematic reduction of the size of the search region // AIChE Journal. – 1973. – Vol. 19, N. 4. – P. 760-766.
18. Бирюков В.Н. Оценка точности определения параметров моделей полевого транзистора // Известия вузов. Электроника. – 2010. – Т. 15, № 4. – С. 22-27.
19. Бирюков В.Н. Обусловленность практических задач параметрической оптимизации // Известия ЮФУ. Технические науки. – 2012. – № 6 (131). – С. 167-169.
20. Пилипенко А.М. Применение гибридной аналитической модели для аппроксимации вольт-амперных характеристик МОП-транзисторов в широком диапазоне температур // Наука и образование: электронное научно-техническое издание. – 2014. – № 8. – С. 268-279.
21. Pilipenko A.M., Biryukov V.N. Modeling of MOSFETs Parameters and Volt-Ampere Characteristics in a Wide Temperature Range for Low Noise Amplifiers Design // Proceedings of IEEE East-West Design & Test Symposium (EWDTS’2014) – Kiev, Ukraine, September 26–29, 2014. – P. 156-159.
22. Anderson E. J., Ferris M.C. A direct search algorithm for optimization with noisy function evaluations // SIAM Journal on Optimization. – 2001. – Vol. 11, N. 3. – P. 837-857.
23. Калиткин H.H., Литвинцева С.П. Градиентный спуск со случайными шагами // Математическое моделирование. – 1997. – Т. 9, № 7. – С. 63-70.
Bottou L. Stochastic Gradient Descent Tricks // Neural Networks: Tricks of the Trade. Second Edition. – Springer, Berlin, 2012. – P. 421-436.
24. Biryukov V.N., Pilipenko A.M. Diagnostics of the Nonlinear Static Models of a Diode // Journal of Communications Technology and Electronics. – 2009. – Vol. 54, N. 5. – P. 577-582.

Comments are closed.