An approach to the assessment of military-oriented aircraft engineering based on neural-like networks

DOI: 10.34759/vst-2021-4-267-280

Аuthors

Zaharov E. N.^*, Usachev D. V.^**

Leonov Moscow Region University of Technology, Korolev, Moscow Region, Russia

*e-mail: motorkom.z@mail.ru
**e-mail: usachev_denis@mail.ru

Abstract

Quality assessment of the military-oriented aircraft engineering (MOAE) samples is being performed by one of the following techniques: complex, differentiated, mixed, integrated as well as by the economic practicality. Each of these methods has its pros and contras.

The complex technique allows assessing the quality level in aggregate, but it does not allow accounting for all meaningful indicators.

The differentiated technique computes simple quality indicators with account for the meaningful ones affecting the quality of the MOAE samples. This method application causes difficulties in the quality level assessment by the large quantity of simple indicators.

The mixed method allows quality assessment of the MOAE sample at large aggregate of the simple, meaningful and generalized indicators. Accounting for the large quantity of indicators requires complex mathematical calculations.

The integral method is applicable for assessing the MOAE operation efficiency. This method application is practical only when total costs of the sample creation, operation and useful effect of the sample operation are determined.

The sample quality assessment technique by the economic effectiveness is applied only when economic assessment is necessary. With this technique application, a large quantity of data on the sample should be necessarily accounted for.

All these techniques are applicable for the assessment of a single-type MOAE samples, namely of the same type and purpose. For assessing diversified MOAE samples quality indices are being employed

A brief analysis of the above listed techniques allows inferring that their application for the MOAE sample is not always practical. It is stipulated by the following reasons:

• The difficulty of reducing a wide nomenclature of indices to the resulting value expressed in a numerical form;

• The absence of the possibility for accounting for the external factors; 

• The absence of the full pattern of the MOAE sample quality.

All these reasons instigate the search for new approaches and techniques of quality assessment accounting for the MOAE sample specifics.

According to the article «Application of analytical methods of open complicated systems for assessing the quality of designs of weapons, military and special equipment», MOAE is an open complicated system. Hence, the most suitable quality assessment technique for the open complicated systems is the technique for express-assessment of the open complicated systems functioning.

With account for the suggested technique and the approach, applied at present, the algorithm for the quality level assessment of the production was developed. The algorithm for the MOAE quality level assessment consists of two basic blocks. The first block is universal, and it is applied for quality level assessment of practically all kinds of products. As applied to the MOAE the first block consists of the following stages:

1. Setting the goals and tasks for the MOAE quality level assessment at all life-cycle stages. The main life-cycle stages are development, production and operation.
2. Defining the quality indicators nomenclature of the MOAE sample under study is a very important stage for its quality assessment. It is necessary to regard for the composition, structure, operation conditions, design specifications specifications and a number of other parameters while defining the quality indicators nomenclature of the MOAE sample.
3. There are six main techniques for defining the values of product quality indicators. They are measuring, registration, calculation, organoleptic, expert and sociological. All these techniques may be employed as applied to the MOAE samples.
4. Quality indicators values determining of the MOAE samples depends on the selected technique, and the tools used by this method.

The second block of the MOAE samples quality level assessment consists of the following stages:

1. The MOAE sample quality formalization represents its expansion into fundamental composite indicators in the form of hierarchical structure. The algorithm distinguishes internal and external formalization. External formalization means the studied object extraction from the external environment. In this particular case, the object of study is the MOAE sample quality indicator. Internal formalization means the MOAE sample quality indicator representation in the form of the hierarchical structure of the indicators, affecting its quality. Let call these indicators factors, since each lower-level indicator in the hierarchical structure affects the upper-level one.
2. Assessment of all factors of the hierarchical structure, as well as those of different physical nature is being performed according to the unified criterion scale, which envisions the factor state assessment on the assumption of the direct assessment principle on the interval from 0 to 1.
3. A neural-like network is being set based on the hierarchical formalization. The neural-like elements of this network and connections formed between them simulate individual factors. Each layer of the neural-like elements simulates factors of one hierarchy level. A neural-like network can work in two basic ways:
• Deterministic, when all neural-like elements operate according to a deterministic option;
• Statistical, when at least one neural-like element operates using simulation by to one of its characteristics. 

4. The initial data for the MOAE sample can be determined on account of the purpose and structure, qualitative and quantitative characteristics of the operation processes, characteristics of external impacts of various physical nature factors, tactical situations options, characteristics and composition of means interacting with the sample, and characteristics of active counteraction means.
5. According to the pre-determined operating option of a neural-like element in the neural-like network, the compliance level of the MOAE sample with the intended objectives is being calculated.
6. If necessary, factor analysis is performed to check correctness and reliability of the resulting operating model of the neural-like network.
7. Decision making on the compliance level of the MOAE sample with the intended objectives (the requirements of tactical and technical tasks or technical conditions) serve as a basis for:
• Preparation and formation of suggestions and conclusions on the possibility of adopting the developed (tested) MOAE samples with putting them into production;
• Assessing the degree of the MOAE sample employing in real combat conditions; 
• the possibility of the MOAE sample employing in various weather conditions./li>

8. Conclusions on the MOAE sample quality level (in conjunction with its purpose) compare the obtained quality indicator either with the basic one or with quality indicators of the foreign samples computed earlier. If the quality indicator appears less to be than the basic one or the foreign sample, suggestions are being elaborated on the indicators (factors) improvement of the first, second, third etc. hierarchical levels.

The suggested approach to assessing the quality level of MOAE sample possesses the following advantages:

• Apprehensible and accessible formalization (structuring) of the object under study;
• A comprehensive assessment of the MOAE samples quality is being performed with account for the external factors of various physical nature;
• The quality level assessment authenticity is being determined by the possibility of employing all available information (deterministic, calculated, expert);

The ability of quick initial data setting and producing the results in real time.

Keywords:

plasma spraying, functional coatings, technological modes of spraying, numerical modeling, shock interaction, specific contacting surface

References

1. Pas’ko T.V., Tarov V.P. Otsenka kachestva tekhnicheskikh sistem (Quality assessment of technical Organizational and methodological bases of activity systems), Tambov, TGTU, 2016, 96 p.
2. Classification of quality level assessment methods. URL:https://intellect.ml/classification-of-quality-level- assessment-methods-7683
3. Kiran D.R. Total Quality Management: Key Concepts and Case Studies. Butterworth-Heinemann, 2016, 582 p.
4. Metody otsenki urovnya kachestva, 2021. URL: https:/ /znaytovar.ru/new540.html
5. Prokhorov Yu.K. Upravlenie kachestvom (Quality management), St. Petersburg, SPbGUITMO, 2007, 144 p.
6. Korobeinikova E.S. Evolvement of quality management systems effectiveness assessment mechanism in aerospace industry. Aerospace MAI Journal, 2019, vol. 26, no. 3, pp. 210-219.
7. Popov S.M. Shchelokov S.A. Perspektivy razvitiya informatsionnykh tekhnologii, 2013, no. 15, pp. 41-45.
8. Demidov B.A., Velichko A.F., Voloshchuk I.V. Sistemno-kontseptual’nye osnovy deyatel’nosti v voenno- tekhnicheskoi oblasti. Kniga 2. Organizatsionno-metodicheskie osnovy deyatel’nosti v voenno-tekhnicheskoi oblasti (System-conceptual foundations of activities in the military-technical field. Book 2. Organizational and methodological bases of activity in the military-technical field). Kiev, Tekhnologicheskii park, 2006, 1152 p.
9. Zakharov E.N., Usachev D.V. Materialy X Ezhegodnoi nauchnoi konferentsii aspirantov MGOTU «Innovatsionnye aspekty sotsial’no-ekonomicheskogo razvitiya regiona» (14 May 2020, science city Korolev, Moscow region), Moscow, Nauchnyi konsul’tant, 2020, pp. 251-258.
10. Sosedova Ya.G., Gerasimov B.I., Sizikin A.Yu. Standartizatsiya i upravlenie kachestvom produktsii (Standardization and product quality management), Tambov, TGTU, 2012, 268 p.
11. Ryapukhin A.V. Innovative technological projects in the domain of aircraft and aerospace engineering quality management. Aerospace MAI Journal, 2017, vol. 24, no. 1, pp. 132-137.
12. Ponomareva S.V., Shishkina G.V., Mozgova G.V. Metrologiya, standartizatsiya i sertifikatsiya (Metrology, standardization and certification), Tambov, GOU VPO TGTU, 2010, 96 p.
13. Sergeeva I. Institut ekonomiki i upravleniya, 2017. URL: http://sergeeva-i.narod.ru/quality/page3.htm
14. Goswami D. The Importance of Quality Assurance in the Aviation Industry, 2020. URL: https:// medium.com/qualitest/the-importance-of-quality-assurance-in-the-aviation-industry-4e68371bdbb6
15. Sakulin S.A., Sokolov D.A., Vykhovanets V.S. Vestnik Irkutskogo gosudarstvennogo tekhnicheskogo universiteta, 2017, vol. 21, no. 9(128), pp. 83–94. DOI: 10.21285/ 1814-3520-2017-9-83-94
16. Usanin S.N. Sovremennoe sostoyanie i puti razvitiya sistemy podgotovki spetsialistov silovykh struktur. Sbornik statei. Perm, Permskii voennyi institut voisk natsional’noi gvardii Rossiiskoi Federatsii, 2016, pp. 295-301.
17. Li Y. Deep reinforcement learning: an overviev. 2017, 150 p. URL: https://arxiv.org/pdf/1810.06339.pdf
18. Dawson C. Applied artificial neural networks. Mdpi AG, 2016, 258 p.
19. Zakharov E.N. Patent RU 2744361 C1, 05.03.2021.
20. Zakharov E.N., Chechkin A.V. Neirokomp’yutery:razrabotka, primenenie, 2008, no. 7, pp. 28-35.
21. Nikonenko S.I, Kadurin A.A., Arkhangel’skaya E.V. Glubokoe obuchenie. Pogruzhenie v mir neironnykh setei (Deep learning. Immersion in the world of neural networks), St. Petersburg, Piter, 2018, 480 p.
22. Kulakov P.A. Izmerenie v sotsiologii (Measurements in sociology), Novosibirsk, NGASU (Sibstrin), 2005, 124 p.