ADAPTIVE ROBUST CONTROL OF A MULTI-STAGE ELASTICALLY DEFORMED ELECTROMECHANICAL OBJECT UNDER UNCERTAINTY

Discusses the problems of constructing an adaptive robust control for a parametrically indefinite multi-stage nonlinear electromechanical object with elastic properties, synthesized on the basis of iterative (step-by-step) methods of backward traversal of the integrator, applicable for nonlinear objects, affinely parameterized with respect to matrices (vectors) of unknown parameters and reduced to a block lower triangular form ... Simplified synthesis procedures are proposed by the integrator bypass method, based on the elimination of cumbersome analytical calculations of «pure» derivatives of virtual step-by-step controls, by replacing them with filtered analogs, which makes it possible to dramatically simplify the process of analytical synthesis and the structure of the controller. It is also proposed to simplify the procedure for synthesizing tuning functions with simultaneous regularization of integral tuning algorithms by the parametric projection method. The results of a computer study of the constructed adaptive robust systems are presented on the example of a detailed mathematical model of a fourdegree manipulator, which takes into account the elastic properties of transmissions and the electromagnetic dynamics of electric actuators. The digital implementation of the constructed adaptive robust systems is carried out on the basis of authorized programs in the MATLAB/Simulink environment.

Authors: Le Hong Quang, V. V. Putov, V. N. Sheludko, A. D. Skakun

Direction: Informatics, Computer Technologies And Control

Keywords: Multi-stage nonlinear electromechanical object with elastic properties, parametric uncertainty, affine parametrization with respect to unknown matrices (vectors), block lower triangular forms, adaptive robust control, iterative (step-by-step) methods of adaptive traversal of the integrator, simplified procedures with replacement of pure derivatives of virtual controls by their filtered analogs, regularization integral algorithms for tuning parametric projection

---

View full article