Kalman Filter-Based Fault-Diagnosis Scheme for DC Motor

Abstract

For engineering systems in industrial applications, fault detection and diagnostics are critical. A considerable number of primary and secondary protection devices are fitted on industrial instruments. However, for the system to operate safely, it must be closely monitored. Many real-world systems are non-linear in nature, and they are influenced by stochastic noises and disturbances. As a result, failure identification of such systems is critical. In this paper, model-based state estimate methodologies are used to enable predictive maintenance and carry out fault detection for a DC motor model. Kalman Filter and Internal-model based Kalman Filter are two model-based estimate techniques examined. Various faults such as bearing fault, inertia fault, short-circuit and open-circuit faults are explored. The results demonstrate that Internal-model based Kalman performs better in fault identification for various fault conditions.