Regularization partial description in evolutionary search for solutions at self-organization

Authors

  • V. F. Irodov State Higher Education Establishment “Pridneprovsk State Academy of Civil Engineering and Architecture”,Department of Heat Engineering and gas supply, Ukraine https://orcid.org/0000-0001-8772-9862
  • R. V. Barsuk State Higher Education Establishment “Pridneprovsk State Academy of Civil Engineering and Architecture”,Department of Heat Engineering and gas supply, Ukraine https://orcid.org/0000-0002-9666-7496

Keywords:

Inductive method of self-organization models of complex systems, evolutionary search, algorithm, unbiasedness criterion, partial description

Abstract

Purpose. The paper deals with self-organizations method developed by A. G. Ivahnenko. For building mathematical model by this method it is necessary experimental data small amount. It is big advantage in complex systems study, including transient modes different processes. For constructing mathematical models by this method, there is full description of object model replaced by several partial descriptions. It well known that this descriptions complexity gives systems more accurate model. Partial description parameters may include nonlinear. There is proposing for determining these parameters using regularization.

Methodology. There are general methods of mathematical models self-organization are gives in the article. Models construction is describe with using parameters that including nonlinear. There is unbiased criterion selected as criterion. Parameters obtaining is depends of input data division. It is propose divide experimental data point by random search. For determining partial description parameters using evolutionary search algorithm. Points random search process displayed in the algorithm.

Findings. There is proposing self-organization method for mathematical method based on experimental data which contain evolutionary search algorithm for determining partial description parameters. Algorithm of this search is constructed. It differs from existing by random division of experimental points. Partial description parameters value is present as mathematical expectation.

Originality. There is proposed new self-organization method for mathematical modeling which build partial description by evolutionary search in regularization process.

Practical value. It is possible modeling quality increasing by this method of constructing mathematical model using. Especially it concerns of complex systems and processes that are more and more expose for modeling.   

Author Biographies

V. F. Irodov, State Higher Education Establishment “Pridneprovsk State Academy of Civil Engineering and Architecture”,Department of Heat Engineering and gas supply

Dr. Sc. (Tech.), Prof.

R. V. Barsuk, State Higher Education Establishment “Pridneprovsk State Academy of Civil Engineering and Architecture”,Department of Heat Engineering and gas supply

Post Grad. St.

References

IvahnenkoA.G., Zaychenko Yu. P.,Dimitrov V. D. Prinyatie resheniy na osnove samoorganizacii. [Making decision based on self-organization] Soviet radio. Moscow. 1976. 128 p

Ivahnenko A. G. Inductivniy metod samoorganizacii modeley slognih system. [The inductive method of self-organization models of complex systems] Naukova Dumka. Kiev. 1982. P.29

IvahnenkoA. G., Muller I. A. Samoorganizaciya prognoziruyushih modeley. [Self-organization of forecasting models] / Technik. Kiev. 1985. 71 p

Irodov V. F. O postroenii i shodimosti algoritmov samoorganizacii sluchaynogo poiska. [The construction and convergence of self-organization algorithms of randoms earch] / Automatic №4. 1987. 34-43 p

Irodov V. F. Regulyarizaciya polinomialnogo pribligeniya funkcii po criteriyu nesmechenosti. [Functions approximation regularization on the criterion of unbiasedness] / Kiev: GRNTB. 1980. №2398. 7 p

Tanana V. P., Belkov S. I. Konechnoraznostnaya approksimaciy ametoda regulyarizacii A. N. Tihonova n-ogo poryadka. [Finite difference approximation of the regularization methods of Tikhonov n-th order] / Vestniko fSUSU. Series of “Computational Mathematics and Computer Science”. Vol.4. №1. 2015 86-98 p

Tikhonov A. N., Arsenin V. Ya. Metodi resheniya nekkorektnih zadach [Methods of solving incorrect problems] / Science. Home edition of Physical and Mathematical literature. 1979. 30 p

Tikhonov A. N. O zadachah s pribligennozadanoy informaciei [Problems with approximately specified information] / Ill-posed problems of nature science. Publishing house of the Moscow University. 1987. 8-14 p

Dianne P. O’Leary. Near-optimal parameters for Tikhonov and other regularization methods / SIAM J. Sci. Comput.Vol.23. № 4.P.1161-1171. Zurich. Switzerland. November 7. 2001

Irodov V. F. Self-organization methods for analysis of nonlinear systems with binary choice relations / System Analysis Modeling Simulation. 1995. V. 18-19. 203-206 p

Garda B. Galias Z. Tikhonov regularization and constrained quadratic programming for magnetic coil design problems / Int. J. Appl. Math. Comput. Sci. Vol. 24. Cracow. Poland. 2014. 249-257 p

Jun He., Mitavskiy Boris, Zhou Yuren. A theoretical assessment of solution quality in evolutionary algorithms for the knapsack problem / Abeystwyth, U. K. Guangzhou, China. 14 Apr. 2014. 15 p

Lemke Frank. Self-organization modeling for decision support / Knowledge Miner Software. Berlin. Germany. ICIM 2013. 172-178 p

Rajkumari Bidyalakshmi Devi, Barlaskar Esha, OinamBinarani Diva, Smriti Priya Medhi, Reingayung RonraShimray. Survey on evolutionary computation tech techniques and its application in different fields / Don Bosco College of Engineering and Technology, Assam Don Bosco University, Guwahati, Assam, India.International Journal on Information Theory.Vol. 3. July 2014. 73-82 p

Published

2015-09-22

Issue

Section

Energy, ecology, computer technology in construction