The condition monitoring system designs potentially-dangerous objects

In this article the approach to classification and formation of informative signs describe the dynamics of processes in potentially dangerous objects, based on the method of normalization of sampling distributions. The regulation allows you to decrease the computational complexity of classification procedures for measurement control devices signals when you build monitoring and control systems. By application shift and large-scale operations studied selective data are provided to an interval [0,1] without structural changes that allows to solve in a new way absolutely many problems of the statistical analysis, including classification of signals.

Important property rated in the range of [0,1] data is opportunity to allocate informative signs of basic selective data for problems of systematization and classification. It is shown that structural properties of selective data are characterized by set of the initial - and the central moments.

Use of a measure of the information - distinctions of selective data and distance between compared signals as criterion of classification allow to simplify significantly classification procedure at creation of monitoring and control systems.