Optimization of decision-making using parametric modeling in a complex system with alternative interconnections between objects

The article focuses on the study of the management decision-making process in an organizational system with alternative supplies, based on parametric optimization modeling. It is shown that this system is characterized by the interconnectedness of the objects in the primary operational environment and the supplier objects, and differences in the performance indicators of the latter lead to diversity in the interaction between the two groups of objects. Selecting the best interaction option necessitates an optimization approach. Since obtaining a management decision based on multi-alternative optimization requires multi-variant selection involving expert information, the result is subjective in nature. For organizing the decision-making process, a more effective optimization modeling method is parametric optimization, which in turn requires preliminary parametric identification of the dependencies of extreme and boundary requirements on the optimized variables, based on retrospective data from digital monitoring. In order to transition to algorithmization of the search for a solution to the constrained optimization problem, a shift is made to an equivalent optimizable function with additional variables that regulate the fulfillment of boundary conditions. A combined algorithm is proposed, which uses an analogue of movement in the direction of the gradient of the equivalent function through random search over a given interval of the optimized variables, as well as per-object gradient search in the space of additional variables, subject to their non-negativity. The result of the search is an optimal variant of supplier characteristics. A comparison of the indicators of potential suppliers with the optimal solution is carried out based on visual-expert evaluation. To this end, the multidimensional space of characteristics is visualized on a plane while preserving the spatial arrangement of points corresponding to the analyzed objects, which makes it possible to identify the supplier whose location is closest to the point representing the coordinates of the optimal solution.

1. Novikov D.A. Theory of management of organizational systems. Moscow: Fizmatlit; 2007. 583 p. (In Russ.).

2. Granin S.S., Mandel A.S. Optimization of inventory management process in a supply chain having alternative. Control Sciences. 2018;(6):24–28. (In Russ.). https://doi.org/10.25728/pu.2018.6.3

3. Weill P., Woerner S. What's Your Digital Business Model?: Six Questions to Help You Build the Next-Generation Enterprise. Moscow: Alpina Publisher; 2019. 257 p. (In Russ.).

4. Lvovich Ya.E., Lvovich I.Ya., Choporov O.N., et al. Optimization of digital management in organizational systems. Voronezh: Nauchnaya kniga; 2021. 191 p. (In Russ.).

5. Shevyreva E.A. Structural modeling of management process in organizational system with alternative supplies. Modeling, Optimization and Information Technology. 2026;14(1). (In Russ.). https://doi.org/10.26102/2310-6018/2026.52.1.011

6. Lvovich I.Ya. Decision-making based on optimization models and expert information. Voronezh: Nauchnaya kniga; 2023. 231 p. (In Russ.).

7. Poroshin I.E., Titov M.A. The use of artificial neural networks for parametric identification of objects. Young scientist. 2020;(26):39–42. (In Russ.).

8. Khnaev O.A., Pchelintsev I.A. Parametric optimization of systems. Methods for solving extreme problems. Models, Systems, Networks in Economics, Engineering, Nature and Society. 2012;(2):146–152. (In Russ.).

9. Korchagin S.G., Ryndin A.A., Ryndin N.A. Management in organizational systems based on digital technologies. Voronezh: Nauchnaya kniga; 2025. 248 p. (In Russ.).

10. Donskoy V.I. Intelligent optimization based on machine learning: state of art and perspectives (A survey). Tavricheskii vestnik informatiki i matematiki. 2020;(1):32–63. (In Russ.).