Unified approach to decentralized control

Group of methods for decentralized control of the interacting multivariable systems is investigated in this paper. Common base of these methods is knowledge of stability limit (ultimate gain and ultimate period) of controlled systems, which can be relatively easily obtained by mathematical solution (when full system model is known) or by experimental way. Controllers tuning may be realized by Ziegler-Nichols method (or by some of its modification) or by trial and error method. All methods are based on solving the system of similar nonlinear equations with various number of unknown parameters.