S.Ya. Grodzenskiy - Dr.Sc. (Eng.), Professor, Moscow University of Technology (MIREA) V.P. Kiselevich - Ph.D. (Chem.), Deputy Director for Production, JSC «Concern «Morinsis-Agat» (Moscow) M.Yu. Khramov - Ph.D. (Eng.), First Deputy General Designer - Director, Scientific-Technical Complex of JSC «Concern «Morinsis-Agat» (Moscow)

Selection and justification of the mathematical model of failure is an important step in the task of forecasting of reliability numbers of electronic equipment. The most common models are the failure models, based on the distribution of the relevant random variables - operating time to failure. In terms of reliability the main types of distributions are divided into the «aging», in which the failure rate function is monotonically increasing, and «youthful» with falling-failure rate. The most realistic model of failure is described by Weibull distribution. However, when the time dependence of the failure rate is expressed by the periods of run-in and wear-out, the Weibull distribution is insufficient. At the article is proposed to use a mixture of exponential and Weibull distributions. In this case, it is possible to give a definite physical meaning and conduct statistical and physical analysis of the electronic equipment reliability for the parameters of mixture distribution.

 

1. GOST R 27.302 -2009. Nadezhnost v tekhnike. Analiz dereva neispravnostejj.
2. Grodzenskijj S.JA. Statisticheskie metody kontrolja i upravlenija kachestvom: Ucheb. posobie. Izd. 2-e. M.: MIREHA. 2011.
3. Grodzenskijj S.JA. Menedzhment kachestva: Ucheb. posobie. M.: Prospekt. 2015.
4. KHramov M.JU. Model processa ehkspluatacii obsluzhivaemykh sistem // Vychislitelnye sistemy realnogo vremeni i cifrovye ustrojjstva. 2015. Vyp. 9. S. 183-191.
5. Guk JU.B., Karpov V.V. Teorija nadezhnosti: ucheb. posobie. SPb: SPGTU. 1982.