Broken Rotor Bar Detection in AC Induction Motors using Cascaded Flux-Current State Observer

This paper proposes a broken rotor bar fault detection method in AC induction motors using a cascaded flux linkage and stator current observer. The proposed observer-based algorithm can enhance a fault detection accuracy by addressing major drawbacks of a widely adopted motor current signature analysis (MCSA) method, such as susceptibility to external environmental factors and low reliability. A Gopinath-style flux observer is implemented to indirectly estimate both stator and rotor flux, and a stator current observer is also added to improve the fault estimation accuracy in both low- and high-speed operation regions. This paper also includes an analytical and dynamic induction motor models with one or multiple broken rotor bars emulating fault conditions. A broken rotor bar fault diagnosis system, namely Fault Diagnosis and Amplifying (FDnA) system, is developed and verified in simulation and it fault detection accuracy is compared with existing fault detection methods.