Failure´s Identification for Electromechanical Systems with Induction Motor

In this paper the new approach for guaranteed state estimation for electromechanical systems with induction motor is implemented for failure´s identification in such systems. Approach is based on matrix comparison systems method, and a comparison with some other estimation methods was realized in package MATLAB. The experimental unit made for practical approbation of identification algorithms is presented with received experimental data. The new approach (matrix comparison systems method) for guaranteed state estimation of an induction motor (IM) and detect faults was presented in [1, 2]. State estimation is based on a discrete-time varying linear model of the IM [3, 4] with uncertainties and perturbations, which belong to known bounded sets with no hypothesis on their distribution inside these sets. The approach uses the measurement results. Recursive and explicit algorithm is presented and illustrated by example with real 1M parameters, realized in program package MATLAB.