Reducing the volume of the test sample of spacecraft elements during control tests

Аuthors

Osipov N. A.^*, Musienko A. S.^**, Merkushev O. A.

Mlitary spaсe Aсademy named after A.F. Mozhaisky, Saint Petersburg, Russia

*e-mail: bayes@mail.ru
**e-mail: vka@mil.ru

Abstract

The article describes a methodology, which combines a priori (additional) information and current control data of tests of elements of space vehicles in order to achieve the goals of selection and obtain a higher assessment of the object being pursued.

Proposed the technique is based on the mathematical apparatus of dynamic Bayesian networks, as well as the basic concepts and relationships of the theory of reliability and technical diagnostics of systems. The initial data is information on the technical condition of the elements of space vehicles during the test control, information on the reliability (structural and logical diagrams, failure rates of elements) of technological equipment, as well as diagnostic models linking the types of technical conditions and diagnostic signs of defective products.

It is proposed to use the selection method in a dynamic Bayesian network to identify discrepancies between products and equipment in the process of monitoring and diagnosing the technical condition of the elements of technological equipment and describing the dynamics of the process.

A posteriori inference allows you to combine heterogeneous initial information and incoming new data to obtain a comprehensive assessment during the process and the state of the process equipment in order to make an informed decision by the specialist to continue or suspend the process, if a defect is detected and take measures to eliminate them .

The advantage of this technique is the ability to take into account heterogeneous a priori information, including the format representation of the subject, and the results of control tests.

The implementation of this approach to control studies of a sample of a limited size is given. The possibility is substantiated of the possibility, based on the results of selective inspection tests, to estimate the predicted value of defective products in the entire batch with a sufficient goal for making a decision on selective control.

The method under consideration combines a priori information and data obtained as a result of tests, which, in the course of comparison, makes it possible to obtain the necessary accuracy to detect product defects.

The proposed method can be used by specialists when carrying out control and system testing operations in order to increase the efficiency of selection and detection of defective products.

Keywords:

prior distribution, control trials, likelihood function, Bayesian estimate, posterior distribution, estimation accuracy, sampling control

References

1. GOST 20911-89. Tekhnicheskaya diagnostika. Terminy i opredeleniya (GOST 20911-89. Technical diagnostics. Terms and definitions), Moscow, Standartinform, 2009, 11 p.

2. GOST R ISO 2859-1-2007. Plany vyborochnogo kontrolya posledovatel'nykh partii na osnove priemlemogo urovnya kachestva. Chast' 1. (GOST R ISO 2859-1-2007. Sampling plans for successive lots based on acceptable quality levels Part.1.), Moscow, Standartinform, 2007, 104 p.

3. GOST R MEK 61124-2016. Statisticheskie metody. Planirovanie i obrabotka rezul'tatov kontrol'nykh ispytanii v sluchae postoyannykh intensivnosti otkazov i parametra potoka otkazov (GOST R IEC 61124-2016. Statistical methods. Planning and processing of proof test results in case of constant failure rate and failure rate parameter), Moscow, Standartinform, 2016, 40 p.

4. GOST R 50779.12-2021. Statisticheskie metody. Statisticheskii kontrol' kachestva. Metody sluchainogo otbora vyborok shtuchnoi produktsii (GOST R 50779.12-2021. Statistical methods. Statistical quality control. Methods of random selection of samples of piece products), Moscow, Standartinform, 2021, 20 p.

5. R 50.1.100 – 2014. Statisticheskie metody. Tri podkhoda k interpretatsii i otsenke neopredelennosti izmerenii (R 50.1.100 - 2014. Statistical methods. Three approaches to the interpretation and evaluation of measurement uncertainty), Moscow, Standartinform, 2015, 32 p.

6. Martin O. Baiesovskii analiz na Python (Bayesian analysis in Python), Moscow, DMK Press, 2020, 340 p.

7. Devidson-Pailon Kemeron. Veroyatnostnoe programmirovanie na Python: baiesovskii vyvod i algoritmy (Probabilistic programming in Python: Bayesian inference and algorithms), Saint Pttersburg, Piter, 2019, 256 p.

8. Kurt Uill. Baiesovskaya statistika: Star Wars, LEGO, rezinovye utochki i mnogoe drugoe (Bayesian stats: Star Wars, LEGO, rubber ducks and more), Saint Pttersburg, Piter, 2021, 304 p.

9. Khei Dzh. Vvedenie v metody baiesovskogo statisticheskogo vyvoda (Introduction to methods of Bayesian statistical inference), Moscow, Finansy i statistika, 1987, 336 p.

10. Kokhenderfer M., Uiler T., Rei K. Algoritmy prinyatiya reshenii (Algorithms for decision making), Moscow, DMK Press, 2023, 684 p.

11. Solov'ev V.A., Lyubinskii V.E., Zhuk E.I. Pilotiruemye polety v kosmos, 2012, no. 2 (4), pp. 44-51.

12. Mikrin E.A., Orlovskii I.V., Bragazin A.F., Uskov A.V. Kosmicheskaya tekhnika i tekhnologii, 2015, no. 4 (11), pp. 58–67.

13. Solov'ev V.A., Lysenko L.N., Lyubinskii V.E. Upravlenie kosmicheskimi poletami (Space flight control), Moscow, Izd-vo MGTU im. N.E. Baumana, 2010, vol. 2, 428 p.

14. Osipov N.A., Dorozhko I.V. Izvestiya vuzov. Priborostroenie, 2013, vol. 56, no. 1, pp. 18–26.

15. Dorozhko I.V., Zakharova E.A., Osipov N.A. Trudy Voenno-kosmicheskoi akademii im. A.F. Mozhaiskogo, 2019, no. 669, pp. 216-223.

16. Dorozhko I.V., Osipov N.A., Ivanov O.A. Trudy MAI, 2020, no. 113. URL: https://trudymai.ru/eng/published.php?ID=118181. DOI: 10.34759/trd-2020-113-14

17. Dorozhko I.V., Gorokhov G.M., Kirillov I.A. Trudy MAI, 2022, no. 125. URL: https://trudymai.ru/eng/published.php?ID=168195. DOI: 10.34759/trd-2022-125-23

18. Dorozhko I.V., Ivanov O.A. Trudy MAI, 2021, no 118. URL: https://trudymai.ru/eng/published.php?ID=158259. DOI: 10.34759/trd-2021-118-19

19. Gusenitsa Ya.N., Dorozhko I.V., Kochanov I.A., Petukhov A.B. Trudy MAI, 2018, no. 98. URL: https://trudymai.ru/eng/published.php?ID=90383

20. Privalov A.E., Dorozhko I.V., Zakharova E.A., Kopeika A.L. Trudy MAI, 2018, no. 103. URL: https://trudymai.ru/eng/published.php?ID=101526

Download