Authors

Algorithm for learning a model for diagnosing a refrigeration system based on workflow parameters with limited statistical information

DOI: 10.17586/1606-4313-2025-24-4-91-98
UDC 681.326

Algorithm for learning a model for diagnosing a refrigeration system based on workflow parameters with limited statistical information

Saidanov Viktor O., Yuri E. Tupitsin, Alexander S. Matyunin

For citation: Saidanov V.O., Tupitsin Yu.E., Matyunin A.S. Algorithm for learning a model for diagnosing a refrigeration system based on workflow parameters with limited statistical information. Journal of International Academy of Refrigeration. 2025. No 4. p. 91-98. DOI: 10.17586/1606-4313-2025-24-4-91-98 (in Russian)

Abstract
The article concerns the simulation of failures of refrigeration systems based on statistics on the behavior of parameters. The well-known approaches to building models for diagnosing complex systems are revealed. The structure of the model assumes the decomposition of the system, taking into account the specifics of its functioning. The model's training is based on a recurrent algorithm that uses a priori statistical data about the system. As a result of the training, reference images of inoperable states of the system are modeled. The training is proposed to be conducted using a trigonometric basis. An orthogonal trigonometric basis with an arbitrary domain of orthogonality is proposed, which is an innovative approach that can be used to improve the convergence of the learning process of diagnostic models. This method is based on the mathematical apparatus of functional analysis and allows you to generate data that can be used to train machine learning models. An example of using the proposed algorithm for learning a model for diagnosing a refrigeration system is given.

Keywords: diagnostic parameters, inoperable state; orthogonal transformation, learning process.

All references

1. Bishop M. Pattern recognition and machine learning. Moscow: Dialectics, 2020. 962 p. (in Russian)

2. Mello Rodrigo. Machine learning: a practical approach to the theory of statistical learning. Springer, 2018. 373 p.DOI: 10.1007/978-3-319-94989-5

3. Loban A.V. Information technology of distributed diagnostics of spacecraft. Moscow-Berlin: Direct-Media, 2015. 146 p. (in Russian)

4. Makhutov N.A., Permyakov V.N., Akhmetkhanov R.S. and others. Diagnostics and monitoring of complex technical systems. Tyumen: TIU, 2017. 632 p. (in Russian)

5. Senchenkov V.I. Models, methods and algorithms of technical condition analysis. Saarbrucken: LAP LAMBERT Academic Publishing House, 2013. 377 p. (in Russian)

6. Chunhui Zhao,Furong Gao. Online fault prediction with relative deviation analysis and vector autoregressive modeling. Chemical Engineering. 2015. Vol. 138. pp. 531-543.DOI: 10.1016/j.ces.2015.08.037

7. Lu G., Zhou Yu., Lu S., Li H. New change point detection system for monitoring equipment.Mechanical systems and signal processing. 2017. Vol. 83. pp. 533-548.DOI: 10.1016/J.YMSSP.2016.06.030

8. Budko P.A., Vinogradenko A.M., Litvinov A.I. Experimental studies of the kinetic method of control and diagnostics of technical means.Mechatronics, automation, control. 2014. No.9. pp. 53-58. (in Russian)

9. Liu U.Yu., Gao K.U., Ye G., Ma R., Han J.G. A new method for diagnosing wind turbine bearing malfunctions based on the integrated LMD expansion. Measurement. 2015. Vol. 74. pp. 70-77.DOI:10.1016/J.MEASUREMENT.2015.06.005

10. Skliros S., Esperon M.M., Fahre A., Jennions I.K. An overview of diagnostic methods for hardware systems based on models and managed by data. Diagnostics. 2019. Vol. 20(1). pp. 3-21.DOI: 10.29354/diag/99603

11. Shi P., Liang K., Han D., Zhang I. A new intelligent method for troubleshooting rotating mechanisms based on deep learning and PSO-SVM. Journal of Vibration Engineering. 2017. Vol. 19(8). pp. 5932-5946.DOI: 10.21595/jve.2017.18380

12. Senchenkov V.I., Matyunin A.S. Increasing the convergence of training procedures in the development of diagnostic models of systems. Devices and systems. Management, monitoring, diagnostics. 2020. No. 8. pp. 18-26.
DOI: 10.25791/pribor.08.2020.1197 (in Russian)

13. Senchenkov V., Absalyamov D., Avsyukevich D. Diagnostics of life support systems with limited statistical data on failures. E3S Web conferences (EECE-2019). Vol. 140. No. 05002. 5 p.

14. Zorich V.A. Mathematical analysis. Part 2. Moscow: ICNMO, 2012. 818 p. (in Russian)

15. Kolmogorov A.N., Fomin S.V. Elements of the theory of functions and functional analysis. Moscow: FIZMATLIT, 2009. 572 p. (in Russian)

16. Senchenkov V.I., Absalyamov D.R., Avsyukevich D.A. Setting a set of diagnostic parameters of a system based on the theory of functional spaces. Proceedings of SPIIRAN. 2019. Vol. 18(4). pp. 949-975.https://doi.org/10.15622/sp.2019.18.4.949-975 (in Russian)

17. Muscat Jo.Functional analysis: an introduction to metric spaces, Hilbert spaces and Banach algebras. Springer, 2014. 420 p.DOI: 10.1007/978-3-319-06728-5

Read the full article