Article

Article title NEURAL NETWORK BASED SIMULATION OPTIMIZATION ALGORITHMS OF COMPLEX SYSTEMS
Authors P.V. Afonin, O.Y. Lamskova
Section SECTION VI. COMPUTER COMPLEXES OF NEW GENERATION AND NEUROCOMPUTERS
Month, Year 12, 2009 @en
Index UDC 004. 421. 6
DOI
Abstract This paper is dedicated to the developing and researching of metamodel-assisted simulation optimization algorithms. The two global optimization algorithms based on neural network approximation metamodels are proposed. The results of the algorithms implementation for three standard test functions as well as for the bank system with several cash registers are presented.

Download PDF

Keywords Complex system; simulation; optimization; approximation; neural network; metamodel.
References 1. Лоу А., Кельтон Д. Имитационное моделирование: Пер. с англ. (3-е изд.) – СПб.: BHV, 2004.
2. Haykin S. Neural Networks – A Comprehensive Foundation. Prentice-Hall, 1994.
3. Jin Y. A comprehensive survey of fitness approximation in evolutionary computation // Soft Computing. – 2005. – Vol. 1, № 9. – P. 3-12.
4. Афонин П.В., Ламскова О.Ю. Алгоритм оптимизации на основе локальных нейросетевых метамоделей в реализации для банковской системы // Сб. научных трудов научной сессии МИФИ-2008. – М.: МИФИ, 2008. – Т. 3. – С. 122-123.
5. Афонин П.В., Ламскова О.Ю. Алгоритмы оптимизации на основе имитационного моделирования и нейросетевых метамоделей // Труды Международных научно-технических конференций «Интеллектуальные системы» (AIS’08) и «Интеллектуальные САПР (CAD’2008). – М.: Физматлит, 2008. – Т. 3. – С. 30-36.
6. Hansen N., Ostermeier A. Completely derandomized self-adaptation in evolution strategies // Evolutionary Computation. – 2001. Vol. 2, № 9. – P. 159-196.

Comments are closed.