ESTIMATION OF ACCURACY OF NEURO-NETWORK METAMODELS OF SURFACE EDDY CURRENT PROBES
DOI:
https://doi.org/10.24025/2306-4412.2.2019.171272Keywords:
metamodel, response surface approximation, neural network, computer experiment plan, metamodel adequacy, metamodel informativity, eddy current probe, excitation coil, eddy current densityAbstract
For the problem of synthesis of an eddy current probe with a homogeneous sensitivity zone, its RBF-metamodels which have high computational efficiency are created. These metamodels can be used to design a probe with a given distribution of the density of eddy currents at testing points of the space located on the surface of the conductive object in the testing area of the probe. The excitation coil of a concentric surface eddy current probe is represented by an actuator which is powered by an alternating current and located above the testing object with constant electrophysical parameters. The obtained metamodels are checked for adequacy and informativeness on a complex of statistical indicators with an objective estimation of their statistical significance. At the approximation of the response surface, the computer experiment plan was used, namely, the multidimensional search space was filled with points generated by means of LPτ-sequences that were evenly located on the response surface. The reproducibility of the review surface is checked using the resulting metamodel throughout the modeling area. A reasonable approximation error has been achieved. The results of numerical experiments show the effectiveness of the use of RBF-metamodel for the response surface approximation. The resulting metamodels have an average model error of 6.76 % for the first subregion, 4.8 % for the second and 4.78 % for the third one. Created metamodel is adequate according to Fisher criterion; informative on the coefficient of determination and significantly reliable. At the same time, it allows to reduce the workload of calculations in the tasks of computer designing of eddy current probes. This opens up new possibilities for the synthesis of surface eddy current probes – both parametric and structural-parametric ones.References
V. Ya. Halchenko, R. V. Trembovetska, and V. V. Tychkov, "Development of excitation structure RBF-metamodels of moving concentric eddy current probe", Electrical engineering & electromechanics, no. 2, pp. 28- 38, 2019.
T. Itaya, K. Ishida, Ya. Kubota, A. Tanaka, and N. Takehira, "Visualization of eddy current distributions for arbitrarily shaped coils parallel to a moving conductor slab", Progress in Electromagnetics Research M., vol. 47, pp. 1-12, 2016.
T. Itaya, K. Ishida, A. Tanaka, N. Takehira, and T. Miki, "Eddy current distribution for a rectangular coil arranged parallel to a moving conductor stab", IET Science, Measurement & Technology, vol. 6, no. 2, pp. 43-51, 2012.
K. Ishida, T. Itaya, A. Tanaka, and N. Takehira, "Magnetic field analysis of an arbitrary shaped coil using shape functions", IEEE Transactions on Magnetics, vol. 45, no. 1, pp. 104-112, 2009.
R. V. Trembovetska, V. Ya. Halchenko, and V. V. Tychkov, "Studying the computational resource demands of mathematical models for moving surface eddy current probes for synthesis problems", Eastern-European Journal of Enterprise Technologies, vol. 95, no. 5/5, pp. 39-46, 2018.
M. A. Chuban, "Approximation of the response surface for the use in the process of parametric synthesis of engineering structures", Vestnik Nats. tehn. un-ta ‟HPI‟. Themat. issue: Transport machine-building: coll. of sci. papers. Kharkov: NTU "HPI", no. 43 (1152), pp. 161-164, 2015 [in Russian].
S. G. Radchenko, "Analysis of methods for modeling complex systems", Matematychni mashyny i systemy, no. 4, pp. 123-127, 2015 [in Russian].
E. Burnaev, M. Panov, D. Kononenko, and I. Konovalenko, "Comparative analysis of optimization procedures based on Gaussian processes". [Online]. Available: http://itas2012.iitp.ru/pdf/1569602385.pdf. Accessed on: Nov. 4, 2015.
E. V. Burnaev, P. D. Erofeev, and P. V. Prikhodko, "Highlighting of the main directions in the approximation problem based on Gaussian processes", Trudyi MFTI, vol. 5, no. 3, pp. 24-35, 2013 [in Russian].
N. V. Queipo, R. T. Haftka, W. Shyy, T. Goel, R. Vaidyanathan, and P. K. Tucker, "Surrogate-based analysis and optimization", Progress in Aerospace Sciences, vol. 41, no. 1, pp. 1-28, 2005.
J. H. Frieman, "Multivariate adaptive regression splines", The Annals of Statistics, vol. 19, no. 1, pp. 1-67, 1991.
V. R. Tselykh, "Multidimensional adaptive regression splines", Mashinnoe obuchenie i analiz dannyih, vol. 1, no. 3, pp. 272-278, 2012 [in Russian].
A. G. Ivakhnenko, Inductive method of selforganization of models of complex systems. Kiev: Nauk. dumka, 1982 [in Russian].
P. V. Afonin, and O. Yu. Lamskova, "Algorithms for optimization of simulation models of complex systems based on neural networks", Izvestiya YUFU: Tehnicheskie nauki, Themat. issue: Intellectual CAD, no. 12 (101), pp. 226-232, 2009 [in Russian].
V. Ya. Halchenko, R. V. Trembovetska, and V. V. Tychkov, "The use of neurocomputing at the stage of metamodels development in the process of optimal surrogate antennas synthesis", Visnyk NTUU «KPI». Seriia Radiotekhnika. Radioaparatobuduvannia, no. 74, pp. 60-72, 2018 [in Ukrainian].
R. V. Trembovetska, V. Ya. Galchenko, and V. V. Tychkov, "Methods of improving the accuracy of neural network metamodels of overhead eddy current converters for surrogate synthesis", in 2-nd Sci. Conf. with Іnternat. Participation Non-Destructive Testing in Context of the Associated Membership of Ukraine in the European Union (NDT-UA 2018) (Poland, Lublin, 15-19 Оct. 2018), Poland: USNDT, 2018, no. 2, рр. 47-49 [in Ukrainian].
Downloads
Published
How to Cite
Issue
Section
URN
License
Copyright (c) 2020 Руслана Володимирівна Трембовецька, Володимир Якович Гальченко, Володимир Володимирович Тичков, Анатолій Вячеславович Сторчак The authors who publish in this journal agree to the following terms:The authors reserve the right to authorship of their work and give the journal the right to first publish this work under the terms of the Creative Commons Attribution License CC BY-NC, which allows other persons to freely distribute published work with a mandatory reference to authors of the original work and the first publication of the work in this journal.
Authors have the right to conclude separate additional agreements for the non-exclusive distribution of the paper in the form in which it was published by this journal (for example, posting work in electronic repository or publishing as part of a monograph), provided that the link to the first publication in this journal is maintained.
The journal policy allows and encourages authors to post on the Internet (for example, in repositories of institutions or on personal websites) the manuscript of work, both before the submission of this manuscript to the editorial staff, and during its editorial work, as it contributes to the emergence of productive scientific discussion and positively affects the efficiency and dynamics of published work citation (see The Effect of Open Access).
