Induction motor tacholess fault detection in transient speeds using adaptive generalized Vold Kalman filter

Fault detection are essential in machine condition monitoring to ensure reliable and efficient operations. This article introduces a methodology to achieve this goal, using tacholess method. Techniques such as the Hilbert transform, adaptive observer, generalized demodulation transform (GDT), and Vold-Kalman filtering (VKF) are combined. The novelty introduced in this article lies in the use of the adaptive observer and the incorporation of the estimated speed to make the generalized Vold-Kalman filter adaptive as a function of motor current. Experimental evaluations were carried out, simulating artificial defect, corresponds to a scenario where the rotor has multiple impacts per revolution, often associated with issues such as misalignment or imbalance. These experimental findings demonstrate the performance of the proposed method, with an adaptive filter showing improved signal extraction across a range of operating situations. A diagnostic protocol has been developed, using statistical analysis, to guarantee the applicability of the method under all operating conditions. • Proposed solutions for transient diagnosis of Induction motor. • Introduction of an adaptive Vold Kalman Filter using estimated motor speed. • Experimental results on emulated default which is difficult to extract from electrical components.