For example,Бобцов

ESTIMATION OF THE MAXIMUM ERROR OF A COMPLEX SYSTEM ACCORDING TO PRIOR AND EXPERIMENTAL DATA

Annotation

The problem of estimation of the maximum error of complex system using the data, received before and after testing of prototypes, is considered. The method of experimented information priority is applied; the method allows to receive estimations, that are most accurate if the prior information does not contradict results of experiments. Analytical expressions are derived for a posterior estimate of the maximum error of complex system and for the gain in accuracy or number of experiments. An example of application of obtained test results is given. The resulting gain is shown to depend on the proximity of a priori and experimental information. The measure to determine the weights for a priori and experimental data is the likelihood ratio for testing the hypothesis of their homogeneity. 

Keywords

Articles in current issue