Abstract
This paper introduces a new physical considerations-based approach to the development of control algorithms for dynamic plants, which involves a compensation principle as follows. Being external with respect to a controlled plant, control actions (compensating signals) are applied with the opposite sign to the corresponding variables of the inverse mathematical model of the plant. Regularization of the resulting system is performed by incorporating etalon filters in compensation loops. The described method is used to solve a control problem for the linear stable multidimensional plants with additive external influences and compensation of the perturbations affecting the output variables.