EVALUATION OF SYSTEM CONTROLLED PARAMETERS INFORMATIONAL IMPORTANCE, TAKING INTO ACCOUNT THE SOURCE DATA INACCURACY

Abstract

The work considers system controlled parameters information value assessing technology in the task of its state identifying. The purpose of the study is to improve the standard methodology for controlled parameters information value assessing. The proposed method is based on the controlled parameters value probabilities analysis falling into the subintervals of the interval of possible values for different states of the system. When the value of the controlled parameter falls into the left or right boundary subintervals of the compatibility interval for any state of the object, the conclusion about its state is made taking into account possible errors of the first or second kind in this case. When the controlled parameter value enters the central subinterval, useful information appears if the corresponding probabilities for the states H1 and H2 are differ significantly. Thus, it is shown that taking into account the probabilities of fuzzy values of the controlled parameter falling into the compatibility interval for various states of the object significantly increases its informational value.