The experimental data for large samples can be represented in the form of a histogram, which is a graphic (empirical) estimate of the density probability. The number of the experimental data group intervals can be defined with the Williams formula. One can build the empirical distribution function based on the empirical density probability.
Weibull-Gnedenko distribution is used for the purpose of diagnostic characters statistical properties description, since such distribution is versatile enough and covers a wide range of the various processes probability characteristics changes cases by means of the parameters variation.
Along with the lognormal distribution this distribution satisfactorily describes details operating time based on fatigue damage, and mean time between failures of bearings. This distribution is used for an assessment of machines parts and components reliability.
Naumenko A.P. Diagnostic features statistical analysis methods // Science, education, business (“Nauka, obrazovanie, biznes”). - 2011. - pp.188-195 Download the publication
You can order the translation by e-mail: a@dynamics.ru
|