In this work, a new general approach to the measurement of material parameters is proposed. The basis of this approach is the joint use of a computer model, a discriminative model, and a Bayesian network to organize the process of restoring material parameters. The purpose of the measuring system is to provide relevant information about the parameters of the material under investigation. We have successfully attempted to use the concept of self-awareness in a system for the automated determination of complex medium properties in real time. This method combines the advantages of both direct experimental research and numerical reconstruction in a virtual model. The work describes in detail the functioning algorithm the support machine and the structure of the Bayesian network. These two elements are crucial to the proposed measuring complex.