Abstract
<jats:p>The paper presents an approach to solving a broad class of logistics problems based on the methods of the theory of optimal set partitioning. In particular, a mathematical for-malization is proposed for problems related to spatial allocation of objects, route opt-imization, territory zoning, and resource distribution, formulated as optimal set partitioning problems according to predefined efficiency criteria. The main focus is on the dynamic asp-ects of the problem formulation, which involve time-varying input data as well as stochastic components and uncertainty. Algorithmic approaches to solving such problems are outlined, including the use of modified heuristic and metaheuristic methods, as well as elements of machine learning for forecasting changes in system parameters. The obtained results can be used to improve the efficiency of decision-making in logistics, urban planning, and other app-lied fields where it is essential to consider both spatial structure and dynamic changes.</jats:p>