CORRECTION OF TRANSFORMATION FUNCTION IN MEASUREMENT DEVICES USING NEURAL NETWORK APPROACH
DOI:
https://doi.org/10.26906/SUNZ.2025.2.041Keywords:
nonlinearity, correction, transformation function, artificial neural network, multilayer perceptron, radial basis neural network, learningAbstract
The article considers ways to reduce the impact of the nonlinearity of the transformation function of measuring instruments on the measurement result accuracy by using an additional corrector device that implements the inverse dependence of the transformation function. The purpose is to study the possibilities of using artificial neural networks as such a corrector, namely a multilayer perceptron and a radial-basis neural network. Using computer simulation modeling, the performance of the proposed methods for correcting the transformation function and the influence of the type of nonlinearity on the quality of such correction were investigated. A comparative analysis with traditional approaches, namely a corrector based on a polynomial approximator was carried out. The modeling results indicate that the accuracy of neural network correctors is not inferior to the accuracy of the polynomial corrector, and in some cases even higher. This opens up prospects for wider application in measuring equipment of such modern methods of processing measurement data as artificial neural networks.Downloads
References
1. Грановский В.А., Сирая Т.Н. Методы обработки экспериментальных данных при измерениях. Л.: Энергоатомиздат, 1990. 288 с.
2. Bishop C.M. Pattern Recognition and Machine Learning. Springer. 2006. 738 p.
3. Haykin S. Neural Networks and Learning Machines. 3 rd. ed. Pearson, 2009. 906 p.
4. Теорія та практика нейронних мереж : навч. посіб. /Л. М. Добровська, І. А. Добровська. К.: НТУУ «КПІ» Вид-во «Політехніка», 2015. 396 c.
5. Hornik K., Stinchcombe M., White H. Multilayer feedforward networks are universal approximators. Neural Networks. 1989. 2(5). pp. 359-366.
6. Запорожец О.В., Коротенко В.А., Овчарова Т.А. Компенсация нелинейности функции преобразования измерительных устройств с помощью искусственной нейронной сети. Системи управління, навігації та зв’язку. Вип. 4(16). 2010. С. 99–103.
7. Дегтярев А.В., Запорожец О.В., Овчарова Т.А. Адаптивная система компенсации нелинейности функции преобразования измерительных устройств на базе трехслойного персептрона. Электротехнические и компьютерные системы. № 06(82). 2012. С. 235–241.
8. Запорожец О.В., Овчарова Т.А. Компенсация нелинейности динамических измерительных преобразователей при помощи нейросетевых моделей. Метрологія та прилади. Вип. № 1/II/(45). 2014. С. 74–77.
9. Запорожец О.В., Овчарова Т.А., Руженцев И.В. Компенсация нелинейности полупроводникового терморезистора с помощью искусственной нейронной сети. Системи обробки інформації. 2015. №6 (131). С. 64–67.
10. Zaporozhets O.V., Shtefan N.V. Using Artificial Neural Network for Compensation of Semiconductor Thermistor Nonlinearity. 2019 IEEE 8th International Conference on Advanced Optoelectronics and Lasers (CAOL), Sozopol, Bulgaria, 6-8 Sept. 2019. PP. 703–706.
Downloads
Published
Issue
Section
License
Copyright (c) 2025 S. Avakin, S. Dovhopolyi, O. Zaporozhets, I. Moshchenko
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
