Measurement and Computing Issues in Control of Complex Systems

The problem of forming standards and procedures for calculating indicators of the functioning of complex systems in conditions of uncertainty is solved. The methodology is based on the methodology of Bayesian intelligent technologies, which makes it possible to carry out modeling in conditions of limited, incomplete and fuzzy data, including single measurements. This methodology includes representation of measurements in the form of a fuzzy number; identification of the type of distribution; generation of samples of a given volume for factors characterizing the result and complex system functioning conditions in a given period of time; construction of a regression model based on the generated samples. The constructed models can used as a tool for calculating dynamic norms by substituting appropriate factors into the equation.