Diagnosing Broken Rotor Bar Faults in Speed-Sensorless Squirrel Cage Induction Motors Using Principal Component Analysis and Knowledge Graph

Abstract

Squirrel cage induction motors (SCIMs) are integral to numerous industrial applications, and the accurate monitoring and diagnosis of rotor bar conditions are paramount for enhancing system productivity and minimizing maintenance expenditures. However, the existing techniques for diagnosing broken rotor bars (BRBs) faults are often constrained by the voltage imbalance conditions and the low slip in practical applications. This paper introduces an innovative approach to BRBs fault diagnosis, using principal component analysis (PCA) and knowledge graph (KG) methodologies. PCA, which is robust to imbalanced conditions, is employed to demodulate the stator current signals and isolate the phase modulation (PM) component. The calculated PM index serves as a fault indicator. Concurrently, the KG framework is introduced to detect the BRB fault and quantify the severity. Experimental results demonstrate the effectiveness and reliability of the proposed method under different load levels and fault severities.