Розробка методичного підходу з оцінки стану складних організаційно-технічних систем

In this study, the object of research is organizational and technical systems. The primary issue addressed is the enhancement of operational efficiency in assessing the state of such systems under constraints related to data reliability, regardless of the volume of incoming data acquired from information sources. The subject of the research is the process of evaluating the condition of organizational and technical systems. The study presents the development of a methodological approach for assessing the condition of complex organizational and technical systems. The originality of the proposed approach lies in the implementation of advanced auxiliary procedures that enable the following: – deployment of a search population of hippopotamus agents across the search plane, accounting for uncertainty in the information acquired through technical means about the organizational and technical system, by employing appropriate corrective coefficients. This enables a reduction in the time required for initial configuration of the subsystem responsible for processing heterogeneous data from extraction sources; – additional consideration of the velocity of each agent within the hippopotamus swarm, allowing the prioritization of search tasks by each individual agent within the corresponding search space (across elements and components of the organizational and technical system); – verification of the algorithm’s convergence to local and global optima; – replacement of ineffective search agents by refreshing the population of hippopotamus agents; – implementation of deep learning mechanisms for the knowledge bases of the hippopotamus agent swarm; – estimation of the required computational resources in cases where the available computational capacity is insufficient for performing the necessary calculations. An illustrative example demonstrated a 13–16 % increase in decision-making efficiency due to the integration of additional procedures, while ensuring a decision reliability level of 0.9