IMPLICIT ESTIMATES FOR THE NB_τ TEST PLAN 

Mikhailov Viktor Sergeevich, lead engineer, Central Research Institute of Chemistry and Mechanics named after D. I. Mendeleev (115487, 16а Nagatinskaya street, Moscow, Russia), Mvs1956@list.ru

621.382.029.6

10.21685/2307-4205-2018-1-8

Abstract. The purpose of this paper is to construct an estimate that is implicitly defined and close in its properties to an effective estimate based on an integral approach for tests conducted in accordance with a plan of type NB τ . 
Methods. To find an effective estimate, we used the integral numerical characteristics of the accuracy of the estimate, namely, the total square of the displacement (deviation) of the expected realization of a certain valuation variant from all possible values of the estimated characteristic from the different values of the parameter of the Poisson crt characterizing the failure flow of the set of products under test.
Results and conclusions. The integral approach has shown its effectiveness in identifying the properties of implicitly given estimates. An implicitly given point estimate of the mean time to failure T is effective among the proposed estimates. As a point estimate of the probability of failure-free operation, a traditional, unbiased and absolutely effective point estimate Pˆ = R / N should always be used, except for fail-safe tests. In this case, an implicit, biased point estimate of the probability of fail-safe operation pˆ should be used.

binomial distribution law, exponential distribution, test plan, point estimation