Fuzzy relational cognitive temporal models for analyzing and state prediction of complex technical systems

The effectiveness of fuzzy cognitive modeling methods for analyzing and predicting the state of complex technical systems (STS) is justified by the following reasons: significant interdependence, non-linear nature and incompleteness of information about the mutual influence of the analyzed parameters of the CTS; a variety of effects of internal and external factors on the CTS; complexity and cost of conducting experimental studies during the operation of these systems. The main limitations of fuzzy cognitive models for modeling STS dynamics are: the complexity of taking into account the mutual influence of parameters with their different time lags relative to each other; the need for their constant operational adjustment and training of component models for all parameters during the operation of the CTS. In this paper, Fuzzy Relational Cognitive Temporal Models (FRCTM) are developed. These models combine the advantages of various types of fuzzy cognitive models, and at the same time neutralize the main limitations of the analysis and prediction of the state of the CTS, which are inherent in the well- known fuzzy cognitive models. The paper also proposes models of system dynamics that take into account the specifics of the FRCTM. We have also developed an approach and implemented a method for calculating fuzzy dependencies in vector-matrix form for dynamic modeling of the CTS. The proposed method makes it possible to solve the problems of increasing the uncertainty of the results and the output of fuzzy values of the FRCTM concepts beyond the ranges of the base sets due to the execution of mass iterative computations. An example of modeling heterogeneous electromechanical systems based on FRCTM is given. The results obtained are the basis for solving a whole range of tasks of analysis, predictive evaluation, modeling of different scenarios of the functioning and development of heterogeneous electromechanical systems for various system factors, operating modes and external conditions.