In this publication a Doctrine for the Conditional Extremization of the Hybrid-Optional Effectiveness Functions Entropy is discussed as a tool for the Reliability Assessments of Engineering Systems. Traditionally, most of the problems having been dealt with in this area relate with the probabilistic problem settings. Regularly, the optimal solutions are obtained through the probability extremizations. It is shown a possibility of the optimal solutions “derivation”, with the help of a model implementing a variational principle which takes into account objectively existing parameters and components of the Markovian process. The presence of an extremum of the objective state probability is observed and determined on the basis of the proposed Doctrine with taking into account the measure of uncertainty of the hybrid-optional effectiveness functions in the view of their entropy. Such approach resembles the well known Jaynes’ Entropy Maximum Principle from theoretical statistical physics adopted in subjective analysis of active systems as the subjective entropy maximum principle postulating the subjective entropy conditional optimization. The developed herewith Doctrine implies objective characteristics of the process rather than subjective individual’s preferences or choices, as well as the states probabilities maximums are being found without solving a system of ordinary linear differential equations of the first order by Erlang corresponding to the graph of the process. Conducted numerical simulation for the proposed mathematical models is illustrated with the plotted diagrams.

Herein it has been made an attempt to find some mathematical models of the main laws of psychology on the basis of the variational principle of the subjective entropy maximum. On the basis of the subjective entropy of an individual preferences extremization principle, using the necessary conditions for extremums of a functional to exist, we get the widely known main fundamental laws of psychophysics: the Bouguer-Weber, Weber-Fechner, Stevens, Zabrodin laws. It has been considered a few special cases of the models for the aggregating functions for sensations and perceptions, as well as cognitive functions for preferences and desires. Also it was obtained expressions for canonical distributions in the case of two-dimensional distribution of sensations and perceptions. The proposed approach, postulating the extremality of human's psych functioning, has the theoretically substantiated value for researches with the use of the principle for the general empirical laws in applied psychology.