Научная электронная библиотека ULRICH'S Periodicals Directory

Сибирский государственный университет
науки и технологий имени академика М.Ф. Решетнева

Сибирский аэрокосмический журнал
ISSN 2712-8970 (Print) ISSN 2782-5760 (on-line)

Сведения об образовательной организации

UDK УДК 504.054 Doi: 10.31772/2587-6066-2018-19-4-574-580
NONPARAMETRIC ALGORITHMS FOR RESTORATION OF RANDOM FIELDS
E. N. Bel’skaya , A. V. Medvedev , E. D. Mikhov , O. V. Taseiko
Reshetnev Siberian State University of Science and Technology, 31, Krasnoyarsky Rabochy Av., Krasnoyarsk, 660037, Russian Federation; Siberian Federal University, 79, Svobodny Av., Krasnoyarsk, 660041, Russian Federation
Numerous practical tasks are closely connected with the need to restore fields of one nature or another from noisy experimental data. A feature of this problem is that a priori information isn't often enough for the description of this field accurately to within a set of a vector of parameters. This is due to the fact that information can be polytypic on the various channels of multidimensional processes. It means that the information matches to various levels of a priori in- formation. In this article special attention is paid to this problem. The corresponding algorithms of identification are used in the presence of a priori information of parametric type. In the presence of a priori information of parametric type, it is advisable to use the appropriate identification algorithms when the structure of field models is defined accu- rately to within a set of parameters and their subsequent evaluation, as current information arrives. If a priori informa- tion isn't enough, it is expedient to the researcher to use nonparametric estimates of Nadaraya–Watson for restoration of the respective fields. At the same time it is essential to determine whether all the channels of the multidimensional system are N- or T-processes. N- or T-processes are processes at which entrance or output components are stochastic – dependent and this dependence is unknown in most cases. The fields of distribution of impurity of harmful substances in atmospheric air of the city are considered as an example of use of similar processes. Nonparametric assessment of function of regression is applied as an algorithm of restoration of this field.
Keywords: nonparametric modeling, local approximation, modeling of an ecological situation.
References


1. Medvedev A. V. Osnovi teorii neparametricheskikh sistem. Identifikatsiya, upravlenie, prinyatiye resheny [Bases of the theory of nonparametric systems. Identification, managements, decision-making]. Krasnoyarsk, Reshetnev University Publ., 2018, 732 p.

2. Medvedev A. V. Osnovi teorii adaptivnykh sistem [Bases of the theory of adaptive systems]. Krasnoyarsk, SibSAU Publ., 2015, 526 p.

3. Bel’skaya E. N. et al. [Environmental assessment using by non-parametric modeling]. Ecology and industry of Russia. 2017, Vol. 8, P. 54–58 (In Russ.).

4. Bel’skaya E. N. et al. [Application of nonparametric modeling in solving problems of environmental monitoring]. Vestnik SibGAU. 2016, Vol. 17, No. 1. P. 10–18 (In Russ.).

5. Bel’skaya E. N. et al. Vosstanovlenie polei zagriazniayshikh veshestv v gorodskoi srede [Reconstruction of air pollution fields in the urban territory]. Reshetnev readings: materials of XX of the International scientific and practical conference. Krasnoyarsk, SibSAU Publ., 2016, Part 2, P. 286–287 (In Russ.).

6. Bel’skaya E. N. et al. Urban environmental assessment with using nuclear approximation]. Safety in a technosphere. 2017, Vol. 6, No. 4, P. 13–20 (In Russ.).

7. Nadaraya E. A. Neparametricheskoe otsenivanie plotnosti veroyatnostei i krivoi regressii. [Nonparametric estimation of density of probabilities and curve regression]. Tbilisi, Tbil. Un-t Publ., 1983, 194 p.

8. Nadaraya E. A. [Nonparametric estimation of curve regression]. Some questions of the theory of probabilistic. 1965, Vol. 5, P. 56–68.

9. Koshkin G. M., Piven I. G. Neparametricheskaya identifikatsiya stokhasticheskikh ob’ektov PDF [Nonparametric identification of stochastic objects PDF]. Khabarovsk, RAN Dal’nevostochnoe otdelenie Publ., 2009, 336 p.

10. Hardle V. Prikladnaya neparametricheskaya regressiya [Applied nonparametric regression]. Moscow, Mir Publ., 1993, 349 p.

11. Zipkin Ya. Osnovy teorii obuchauschikhsya sistem [Bases of the theory of the trained systems]. Moscow, Nauka Publ., 1970, 252 p.

12. Vasilyev V. A., Dobrovidov A. V., Cat’s G. M. Neparametricheskoe otsenivanie funktsionalov ot raspredelenyi statsionarnykh posledovatelnostei [Nonparametric estimation of functionalities from distributions of the stationary sequences]. Moscow, Nauka Publ., 2004, 508 p.

13. Komissarov Ju. A. et al. Ekologichesky monitoring okruzhaushey sredi [Ecological monitoring of the environment]. Vol. 1. Moscow, Khimiya Publ., 2005, 365 p.

14. Antropov K. M., Kazmer Y. I., Varaksin A. N. [Description of spatial distribution of pollution of atmospheric air of the industrial center by the Land Use Regression method (review)]. Ekologicheskie sistemy i pribory. 2010, No. 1, P. 28–41 (In Russ.).

15. Informatsionnaya sistema dlya modelirovania rasprostranenia zagryazneniya atmosfernogo vozdukha s ispolzovaniem ArcGIS [Young scientist. Information system for the modeling of air pollution using ArcGIS] (In Russ.). Available at: http://www.moluch.ru/conf/tech/archive/4/895/ (accessed: 14.07.2018).




Bel’skaya Ekaterina Nikolaevna – Cand. Sc., Docent, Department of Health and safety, Reshetnev Siberian State

University of Science and Technology. Е-mail: ketrin_nii@mail.ru.

Medvedev Alexandr Vasil’evich – Dr. Sc., professor, Department of System analysis and research of operations,

Reshetnev Siberian State University of Science and Technology.

Mikhov Evgenii Dmitrievich – head of the methodics department, Military training center, Military engineering

institute, Siberian Federal University. Е-mail: edmihovi@mail.ru.

Taseiko Olga Viktorovna – Cand. Sc., Docent, head of Department of Health and safety, Reshetnev Siberian State

University of Science and Technology. Е-mail: taseiko@gmail.com.


  NONPARAMETRIC ALGORITHMS FOR RESTORATION OF RANDOM FIELDS