V.M. Antoshina – Part-programming Engineer, 

JSC «A.L. Mints Radiotechnical Institute» (Moscow), HSSI MPTI

E-mail: vantoshina@rti-mints.ru

I.B. Zager – Head of the Educational Part – Deputy Head of the Department, 

Lomonosov Moscow State University

E-mail: igor2007@bk.ru

A.S. Logovsky – General Designer, 

JSC «A.L. Mints Radiotechnical Institute» (Moscow)

E-mail: logovsky@rti-mints.ru

K.V. Lvov – Student, 

Faculty of Physics, Lomonosov Moscow State University

E-mail: lvov.kv14@physics.msu.ru

A.Yu. Perlov – Head of Sector, 

JSC «A.L. Mints Radiotechnical Institute» (Moscow) E-mail: aperlov@rti-mints.ru

The paper considers a simulation model of data on the technical condition of the functional systems of radar stations in solving the problem of forecasting failures. Analysis of the time spent on sampling for high-precision prediction of station failures showed that the performance of work by known methods by testing equipment at the stands does not allow us to provide the required time for the creation of a radar station with an automated operation control system.

The existing level of product unification allows using both data from the previous generation of stations and current data from the built-in monitoring of new generation radar stations, due to their correlation relationships. These features create the conditions for the rapid formation of a large array of data on the technical condition of the stations due to the development of a simulation data model of built-in control. A feature of such a model should be its versatility, i.e. the formation of the data stream of both binary sensors (operational or not), and sensors of physical parameters.

To determine the minimum required sample size, an analysis was made of the temperature values recorded during the testing of the components of the product transmitting system. To determine the effect of the training sample volume on the forecast accuracy, the temperature forecast model described above was trained on samples of various volumes. It is shown that for the timely and highquality completion of the development of a radar equipped with an automated operation control system, it is fundamentally important to solve the problem of technical condition data, the processing of which will provide the specified forecast accuracy.

The developed methodological apparatus for generating data of the station’s built-in control provides a solution to the problem of high-precision failure prediction. The approach proposed in the article generally corresponds to the modern directions of development of the leading military-industrial companies of the USA and Europe.

1. Shmueli G., Koppius O.R. Predictive analytics in information systems research. MIS Quarterly. 2011. P. 553−572.
2. Wang Z. Failure prediction using machine learning and time series in optical network. Optics Express. 2017. V. 25. № 8. P. 18553−18565; Freund Y., Mason L. The alternating decision tree learning algorithm. ICML. 1999. V. 99. P. 124−133.
3. Boev S.F. Upravlenie riskami proektirovaniya i sozdaniya radiolokatsionnykh stantsiy dal’nego obnaruzheniya. M.: Izd-vo MGTU im. N.E. Baumana. 2017. (In Russian).
4. Boev S.F., Logovskiy A.S. Upravlenie protsessami sozdaniya RLS DO funktsional’no-blochnoy struktury. Zhurnal radioelektroniki. 2017. № 7. (In Russian).
5. Romanov A.A. Prikladnoy sistemnyy inzhiniring. M.: Fizmatlit. 2015. 555 s. (In Russian).
6. URL = https://www.raytheon.com/sites/default/files/technologytoday/2018/issue1/wp-content/uploads/2018/08/Raytheon_Technology Today_Issue1_2018.pdf.
7. Sun W.G. et al. Design of Autonomic Logistics Information System Based on System Integration Model. Advanced Materials Research. 2012. V. 472−475. P. 2650−2654.