Macrosystems are systems in which the stochastic behaviour of the elements is transformed into deterministic behaviour of the system as a whole. This publication discusses equilibrium in these systems. Mathematical models of stationary states using the principle of maximum entropy are presented. This is developed and generalized for macrosystems with constrained resources. Parametric properties which characterize a model's response to data variations are discussed. The author has developed new computational methods for the computer-aided realization of stationary state models. Algorithms and computer experiments are evaluated. Mathematical modelling methods are applied to problems of hierarchical structures, interregional product exchange and image reconstruction.