The synthesis of algorithms for parameters estimation of adaptive systems

The problem of current determining the parameters of the adaptive dynamic systems by minimizing the residuals of observations. The developed identification method allows to obtain the new structures of feedbacks with respect to the state variables and the parameters by constructing invariants. They allow to formalize a priori uncertainty in the parameters dynamics through the use of additional properties that are exposed at the true trajectory in accordance with the variation method, which follows from the Hamilton - Ostrogradskii principle which is the base for the dynamics of the adaptive system and the parameters. The variations properties allow to decompose the identification system into the subsystems of the parameter and status estimation. The relationship between them is determined by the equation for the sensitivity function. The constructiveness of the synthesized algorithms of identification is confirmed by the results of solution the problem of identifying the parameters of the nonlinear dynamic system regulator.