New Diagnostic and Severity Estimation Method for Inter-Turn Short Fault for Dual Star Permanent Magnet Synchronous Generator

• Published in: Arabian Journal for Science and Engineering Vol. 47; no. 3; pp. 3573 - 3581
• Main Authors: Amirouche, Elyazid, Iffouzar, Koussaila, Houari, Azeddine, Ghedamsi, Kaci, Aouzellag, Djamal
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.03.2022; Springer Nature B.V; King Fahd University of Petroleum and Minerals - Springer
• Subjects: Algorithms; Amplitudes; Circuits; Conductors; Content analysis; Converters; Engineering; Engineering Sciences; Humanities and Social Sciences; multidisciplinary; Permanent magnets; Research Article-Electrical Engineering; Robustness; Science; Short circuit currents; Signal processing; Synchronous machines; Dual star permanent magnet synchronous machine; Diagnostic; Fault severity estimation; Inter-turn short fault
• ISSN: 2193-567X; 1319-8025; 2191-4281
• DOI: 10.1007/s13369-021-06445-2

The dual-star permanent magnet synchronous machine (DSPMSM) is widely used in high power applications where the continuity of service is highly recommended. As they mostly work in harsh environments, they are commonly subject to many types of faults. One of the most dangerous is the inter-turn short fault, which induces torque vibrations and overheats inside the machine, causing the propagation of the fault to nearby conductors and may also induce other faults, like phase-to-phase and phase-to-ground short-circuits. To correctly handle this situation, a proper diagnostic algorithm able to estimate its severity must be used in the control scheme. In this scope, a robust diagnostic and severity estimation algorithm is presented in this paper, based on the analysis of the amplitude and displacement angle of a new defined variable, called ITSF Fault Descriptor (ITSF-FD), without need for harmonic content analysis, making this technic simpler and more robust compared to other published technics which are especially focused on harmonic content analysis and complex signal processing, which are subject to various disturbances, mainly caused by the converter switching and machine’s asymmetries, and they usually need pre-set data or laboratory prototype to extract a threshold specific to each machine. The new variable is defined using measurable machine quantities and reflects the fault characteristics, such as the fault ratio and resistance. It is a sinusoidal signal, its amplitude and displacement angle are mainly influenced by the short-circuit current and are independent of the machine. Simulation results show that the algorithm is capable of diagnosing the fault without need to pre-set data.