DREM Based DC Motor Components Fault Detection and Isolation

The paper is devoted to the application of dynamic regressor extension and mixing approach (DREM) to direct current (DC) motor component fault detection and isolation. Cases of only current sensor, only velocity sensor and both sensors equipment are considered. Linear filters are used to provide model parametrization to autoregression form. Proposed solution allows to detect all component faults and isolate the following components single faults: inductance, resistance, flux and electromechanical constant, inertia momentum, back-EMF constant, friction constant. Simulation results confirm performance of the proposed approach.