Stable Estimation of Regulator Parameters in Direct Adaptive Control Systems for Dynamic Objects

The article offers computational procedures for the formation and construction of algorithms for stable estimation of parameters of regulators in systems of direct adaptive control of dynamic objects. An important advantage of the direct approach is the fact that in systems, the goal of management and the goal of adaptation coincide and consist in ensuring the proximity of the behavior of the system to a given reference model that reflects the idea of the desired behavior of the system. In the proposed adaptive system, the control device contains a predictor of the state of the controlled process, as well as an evaluator and the adaptive controller itself. The system evaluates the parameters of the optimal regulator. When solving the considered incorrect problem, retrospective computational algorithms of simplified regularization are used. The developed algorithms make it possible to find optimal values of the parameters of the controller settings in systems of direct adaptive control of dynamic objects under conditions of a priori statistical uncertainty regarding the statistical characteristics of disturbing influences and interference.