Application of Neural Networks in the Problem of Quantitative Analysis of Air Composition

Purpose of reseach is to develop a method for generating training data to enable the use of artificial neural networks (ANN) method in gas analyzer systems. The problem of increasing the accuracy of separate determination of gas concentrations in multicomponent mixtures under conditions of environmental parameters changes is considered. It is proposed to increase the accuracy of determining target gas concentrations by using the ANN method for joint processing of sensor signals.

Methods: Training data for the neural network were generated using numerical experiments and mathematical simulation methods. To assess the accuracy of training, the standard deviation (SD) was used and the relative error was calculated. ANN training and research were conducted in the MATLAB environment (the Neural Networks Toolbox application). When developing mathematical models of gas sensors, the theory of electrical circuits, electronic theory of chemisorption and the adsorption theory of heterogeneous catalysis were applied.

Results: A method for generating training data sets using mathematical models of gas sensors is described. The proposed training method has been tested on a specific task, in particular, a decision-making device based on ANN for a four-component gas analyzer has been developed. The efficiency of using neural networks for tuning out from the mutual cross-sensitivity of sensors was evaluated.

Conclusion: A method for generating training data using simulation models is proposed, which allows automazing the process of training, research, choosing the architecture and structure of ANN and their testing. The method was tested. Based on the analysis of the obtained errors, conclusions are made about the efficiency of using neural networks to reduce errors caused by cross sensitivity at different concentrations of the main and interfering gases.

1. George F., Cavanagh L. M., Afonja A., Binions R. Metal Oxide Semi-Conductor Gas Sensors in Environmental Monitoring. Sensors, 2010, pp. 5469-5502.

2. Abderrahim H., Berrebia M., Hamou A., Kherief H., Zanoun Y., Zenata K. Measure of carbon dioxide using a gas sensor of a semiconductor type based on tin dioxide (SnO2 ). J. Mater. Environ. Sci., 2011, 2 (2), pp. 94-103.

3. Belysheva T.V., Bogovtseva L.P., Gutman E.E. Primenenie metallooksidnykh poluprovodnikovykh geterosistem dlya gazovogo analiza [The use of metal oxide semiconductor heterosystems for gas analysis]. International Scientific Journal: for Alternative Energyand Ecology, 2004, no. 2, pp.67-66 (In Russ.).

4. Dreizin V., Brezhneva E. O., Bondar O. G. Ustroistvo obrabotki signalov mnogokomponentnogo gazo- analizatora [The signal processing device of a multicomponent gas analyzer]. Pribory i Sistemy. Upravlenie, kontrol', diagnostika = Devices and Systems. Management, control, diagnostics, 2011, no. 12, pp. 43-48 (In Russ.).

5. Brezhnev E.O., Bondar O.G. Gazoanalizator monooksida ugleroda i vodoroda [Gas analyzer of carbon monoxide and hydrogen]. Izvestiya Yugo-Zapadnogo gosudarstvennogo universiteta = Proceedings of the Southwest State University, 2012, no. 2-3, pp. 67-70 (In Russ.).

6. Tomakova R.A., Filist S.A., Yaa Z. Do. Universal'nye setevye modeli dlya zadach klassifikatsii biomeditsinskikh dannykh [Universal Network Models for Classification of Biomedical Data]. Izvestiya Yugo-Zapadnogo gosudarstvennogo universiteta = Proceedings of the Southwest State University, 2012, no. 4-2 (43), pp. 44-50 (In Russ.).

7. Dashkovsky A.A., Primsky V.F. Matematicheskoe modelirovanie mnogokomponentnykh gazoanaliticheskikh izmerenii i analiz pogreshnostei [Mathematical modeling of multicomponent gas analytical measurements and error analysis]. Vostochno-Evropeiskii zhurnal peredovykh tekhnologii = East European Journal of Advanced Technology, 2005, no. 6/2 (18), pp. 108-111 (In Russ.).

8. Dreizin V.E., Brezhnev E. O. [Modeling of gas-sensitive sensors]. Materialy 2-oi Mezhdunarodnoi nauch.-tekhn. konf. "Informatsionno-izmeritel'nye diagnosticheskie i upravlyayushchie sistemy. Diagnostika-2011" [Materials of the 2nd International scientific and technical. conf. "Information-measuring diagnostic and control systems. Diagnostics-2011"]. Kursk, 2011, pp. 53-59 (In Russ.).

9. Bondar O. G., Brezhneva E. O., Pozdnyakov V. V. Methods and Algorithms for Control of a Thermocatalytic Hydrogen Sensor. Measurement Techniques, August, 2018, vol.61, no. 5, pp.514-519.

10. Manginell R. P., Smith J. H., Ricco A. J., Hughes R. C., Moreno D. J. P. Moreno Electro-thermal modeling of a microbridge gas sensor. Sandia National Laboratories, Albuguergue, NM87185-1080, 1997, pp. 360-371.

11. Dreizin V. E., Brezhnev E.O., Bondar O. G. Modelirovanie kataliticheskogo datchika vodoroda [Modeling a catalytic hydrogen sensor]. Izvestiya Yugo-Zapadnogo gosudarstvennogo universiteta = Proceedings of the Southwest State University, 2011, no. 5 (38), pt.1, pp. 69-76 (In Russ.).

12. Drazdale D. Vvedenie v dinamiku pozharov [Introduction to Fire Dynamics]. Moscow, Stroyizdat Publ., 1990, 424 p. (In Russ.).

13. Scheglov P.P., Sharovarnikov A.F. Toksichnye produkty termicheskogo razlozheniya i goreniya polimernykh materialov pri pozhare [Toxic products of thermal decomposition and combustion of polymeric materials in case of fire]. Moscow, 1992, 80 p. (In Russ.).

14. Koshmarov Yu. A. Prognozirovanie opasnykh faktorov pozhara v pomeshchenii [Prediction of hazardous factors of fire in the room]. Moscow, 2000, 118 p. (In Russ.).

15. Dreizin V.E., Brezhneva E.O. Sravnitel'nyi analiz kharakteristik promysh-lennykh gazochuvstvitel'nykh datchikov [Comparative analysis of the characteristics of industrial gas-sensitive sensors]. Datchiki i sistemy = Sensors and systems, 2011, no 3, pp. 68-78 (In Russ.).

16. Sazonov S.Yu., Titenko E.A., Hanis N.A. Podkhod k prognozirovaniyu vozniknoveniya pozharoopasnoi situatsii v data-tsentre na osnove neironnykh setei [Approach to prediction of fire management in the data center based on neural networks]. Izvestiya YugoZapadnogo gosudarstvennogo universiteta. Seriya: Upravlenie, vychislitel'naya tekhnika, informatika. Meditsinskoe priborostroenie = Proceedings of the Southwest State University. Series: Control, Computing Engineering, Information Science. Medical Instruments Engineering, 2015, no. 4 (17), pp. 8-14 (In Russ.).

17. Swingler K. Applying neural networks: A practical guide. London, Academic Press, 1996, 345 р.

18. Foresee F.D., Hagan M.T. Gaus-Newton approximation to Bayesian regularization. Proceedings of the International Joint Conference on Neural Networks, 1997, pp. 1930-1935.

19. Dreyzin V. E., Grimov A. A. Izmeritel'nyi blok dlya neitronnogo spektrometra real'nogo vremeni s vychislitel'nym vosstanovleniem energeticheskikh spektrov s pomoshch'yu neironnykh setei [A Measuring unit for a real-time neutron spectrometer with computational reconstruction of energy spectra using neural networks]. Izvestiya YugoZapadnogo gosudarstvennogo universiteta = Proceedings of the Southwest State University, 2012, no. 2-3, pp. 223-228 (In Russ.).

20. Mikhalchenko D.I., Ivin A.G., Sivchenko O.Yu., Aksamentov Е.А. [Application of Deep Neural Networks in the Problem of Obtaining Depth Maps from Two-Dimensional Images]. Izvestiya Yugo-Zapadnogo gosudarstvennogo universiteta = Proceedings of the Southwest State University. 2019, 23(3): 113-134 (In Russ.). https://doi.org/10.21869/2223-1560-2019-23-3-113-134.