Dr. Alexey Vedyakov, Ph.D.
Faculty of Control Systems and Robotics, ITMO University (Russia)
One of the central problems in control and systems theory, that has attracted the attention of many researchers for several years is the design of observers for the systems state. The problem is particularly challenging if the system is nonlinear and some of its parameters are unknown. This task requires the development of adaptive observers, that should estimate simultaneously the parameters and the systems state.
Usually, the gradient descent-based approach to state observation and parameters estimation is used. However, in the multidimensional case, there is the limited capability to control performance and check online the necessary and sufficient persistent excitation condition. To solve these problems, dynamic regressor extension and mixing method (DREM) for parameters estimation of linear regression models was proposed. Recently a novel adaptive state observer for a class of nonlinear systems with unknown parameters was designed based on DREM.
In the seminar, the problems of online parameters estimation and states observation will be considered. The DREM will be compared with the standard gradient descent approach and least-squares estimators. The adaptive state observers design will be illustrated by examples for electromechanical systems: permanent magnet synchronous motor, induction motor, magnetic levitation system.Bio.
Email Address: firstname.lastname@example.org