Algorithm of real-time developing a forecast model of engineering systems

The article investigates the technical process observability problem that lies in forecasting functions and parameters of a system as a whole or its separate subsystems at a definite operating period and real-time, i.e. the article considers the output forecast construction by the observed input test data. To that end, we designed and investigated the algorithm of choosing the trajectory for developing a forecast function of engineering systems for real-world operation conditions, the algorithm based on aposterior information. Methods and subjects include methods of examining the analytical models of engineering systems, simulation modelling, algorithm theory, implementation of problem-oriented programs for carrying out a simulation experiment, modern technologies of math modelling and simulation experiment. Making comparison between mechanical and electromagnetic oscillation, we offer to investigate one of the options of determining the base (basis) of a forecast function of technical processes in which it is possible to determine the basis and refining indices capable of describing the objects and connections of a real-world system with a formal language at a sequential processing of the test data. As a result, we obtain a functional relationship that makes it possible to forecast the engineering systems behavior, where the forecast accuracy depends on the number of variables used by the given forecast function, i.e. the higher degree of coincidence is required, the more parameters are necessary to be involved.