DOI: 10.18503/1995-2732-2026-24-1-157-167
Abstract
The relevance of this research is determined by the need for domestic manufacturers to find ways to extend the service life of equipment in the mechanical engineering industry. This situation necessitates the development of tools that enable the assessment of the technical condition of equipment using reliability theory methods. Since equipment represents a complex technical system whose components are characterized by different causes of failure, it is advisable to use a superposition of several reliability functions. The aim of the study is to develop and test a program that allows primary data preprocessing, calculation of the parameters for six mixtures of two distributions, selection of the most suitable reliability function, and visualisation of empirical and theoretical distribution graphs. The study uses methods of reliability theory specifically reliability functions that follow three unimodal distributions: exponential, normal, and Weibull. When conducting exploratory data analysis, the Grubbs’ test is used to identify outliers, and the least squares method is applied to select the most appropriate mixture distributions by minimizing the approximation error. The developed programme is written in the Python programming language using standard libraries and modules for data analysis. The novelty of the research lies in the development of a comprehensive software product that allows for a full cycle of data analysis: from the creation of new features to the visualisation of research results. The proposed tool allows comparison of both homogeneous mixtures of two distributions and heterogeneous mixtures using the residual error value, which expands the class of models considered in reliability problems. The result of the program's operation is the selection of the most suitable reliability function, which can be used to predict the time of uninterrupted operation of equipment for a given probability of failure. The developed program represents a ready-to-use tool that can be useful for engineers and reliability specialists to adequately assess the technical condition of equipment.
Keywords
reliability theory, reliability function, superposition of distributions, least squares method, Grubbs’ test
For citation
Novikova T.A., Lyapuntsova E.V. Development of a Tool for Predicting Equipment Technical Condition of Based on the Superposition of Reliability Functions. Vestnik Magnitogorskogo Gosudarstvennogo Tekhnicheskogo Universiteta im. G.I. Nosova [Vestnik of Nosov Magnitogorsk State Technical University]. 2026, vol. 24, no. 1, pp. 157-167. https://doi.org/10.18503/1995-2732-2026-24-1-157-167
1. Survey data. Monitoring predpriyatiy [Enterprises monitoring]. Available at: http://www.cbr.ru/Content/Document/File/135603/mp_survey_data.xlsx. (Accessed on December 08, 2025).
2. Federal State Statistics Service. Available at: https://rosstat.gov.ru. (Accessed on December 08, 2025).
3. Litvinenko R.S., Pavlov P.P., Aukhadeev A.E. Practical application of the normal distribution law in the theory of reliability of technical systems. Izvestiya vysshikh uchebnykh zavedeniy. Elektromekhanika [News of higher educational institutions. Electromechanics]. 2016;(4):96-99. (In Russ.)
4. Litvinenko R.S., Idiyatullin R.G., Aukhadeev A.E. Analysis of the use of exponential distribution in the theory of reliability of technical systems. Trudy mezhdunarodnogo simpoziuma «Nadezhnost i kachestvo» [Reliability and Quality. Proceedings of the International Symposium]. 2016;1:35-38. (In Russ.)
5. Sluchaninov N.N., Fedorov I.V., Koshkin S.V. Determination of the main reliability indicators of functional and structural schemes of mechanisms and machines. Spetsialnaya tekhnika i tekhnologii transporta [Special equipment and transport technologies]. 2020;(6(44)):161-167. (In Russ.)
6. Oboskalov V.P., Kokin S.E., Kirpikova I.L. Primenenie veroyatnostno-statisticheskikh metodov i teorii grafov v elektroenergetike: ucheb. posobie [Application of probabilistic statistical methods and graph theory in the electric power industry: textbook]. Yekaterinburg: Publishing house of UrFU, 2016, 271 p. (In Russ.)
7. Veleulov Z.A., Kinzhagulov I.Yu., Fedorov A.V., Firyulin D.R. Nadezhnost izdeliy i sistem [Reliability of products and systems]. Saint Petersburg: Universitet ITMO, 2023, 184 p. (In Russ.)
8. Babaev S.G., Gabibov I.A., Melikov R.X. Osnovy teorii nadezhnosti neftepromyslovogo oborudovaniya. Uchebnik [Fundamentals of the theory of reliability of oilfield equipment. Textbook]. Baku: AGNA Publishing house, 2015, 400 p. (In Russ.)
9. Kubarev A.I. Nadezhnost v mashinostroenii [Reliability in mechanical engineering]. Standards publishing house, 1977, 264 p. (In Russ.)
10. Stavrovskiy M.E. Osnovy funktsionirovaniya sistem servisa: uchebnik dlya vuzov [Fundamentals of the functioning of service systems: textbook for universities]. Moscow: Yurait Publishing house, 2025, 190 p. (In Russ.)
11. Artyushenko V.M., Bruskov A.A. Application of the Weibull distribution for spacecraft reliability estimates. Vestnik Voronezhskogo gosudarstvennogo universiteta. Seriya: Sistemniy analiz i informatsionnye tekhnologii [Bulletin of Voronezh State University. Series: System analysis and Information Technology]. 2024;(2):89-102. (In Russ.)
12. Sadykhov G.S., Kudryavtseva S.S. Calculation and evaluation of reliability indicators of non-recoverable objects at operating times distributed as a mixture of two exponential distributions. Trudy mezhdunarodnogo simpoziuma «Nadezhnost i kachestvo» [Reliability and Quality. Proceedings of the International Symposium]. 2020;1:29-31. (In Russ.)
13. Fedotova I.M., Vainshtein V.I., Tsibulskiy G.M., Vainshtein Yu.V. The recovery function for operating time with distributions as a mixture of n exponential distributions. Finding the parameters of mixtures by the method of moments. Problemy mashinostroeniya i nadezhnosti mashin [Problems of mechanical engineering and machine reliability]. 2019;(3):104-112. (In Russ.)
14. Efremov L.V. Praktika veroyatnostnogo analiza nadezhnosti tekhniki s primeneniem kompyuternykh tekhnologiy [The practice of probabilistic reliability analysis using computer technology]. Saint Petersburg: Nauka, 2008, 216 p. (In Russ.)
15. Usacheva A.B., Galiaskarov D.R. Python libraries for data analysis. Colloquium-Journal [Colloquium-Journal]. 2020;(29-1(81)):58-59. (In Russ.)
16. Larin S.E., Belash V.Yu. Python libraries for data analysis: data preprocessing and preparation. Dnevnik nauki [Diary of science]. 2023;(12(84)). (In Russ.)
17. Shiryaeva L.K. On some properties of the distribution of two-way Grubbs statistics. Rossiiskaya nauka: aktualnye issledovaniya i razrabotki: Sbornik nauchnykh statey XI Vserossiiskoy nauchno-prakticheskoy konferentsii [Russian science: current research and development. Collection of scientific articles of the XI All-Russian Scientific and Practical Conference]. Samara: Samara State University of Economics, 2021, pp. 43-47. (In Russ.)