A Novel Cost‐Effective Fault Diagnosis Method for Incipient Detection of Broken Rotor Bar in Induction Motor

This paper explores a novel and low‐cost framework for diagnosing incipient broken rotor bar (BRB) faults in a squirrel‐cage induction motor operating under different BRBs and varying load scenarios. First, an adaptive Principal Component Analysis (APCA) is applied to the three‐phase stator currents, isolating the weak fault‐related component from noise and other outlier components. After generating an analytic envelope signal, the fault frequency is estimated from the envelope via spectral analysis using the Goertzel algorithm. Secondly, the rotor speed frequency is estimated from the APCA‐transformed signal to justify the fault frequency extracted through APCA. The estimated rotor speed frequency is then used to calculate the model‐based fault frequency using the known BRB fault equation. Comparing the experimental and theoretical fault frequencies indicates the physical consistency of the signal, thus enhancing diagnostic confidence and minimizing the risk of false positives. The method is validated on both public and laboratory datasets. The results demonstrate that the proposed framework offers an accurate, robust, and sensorless implementation for industrial deployment. © 2026 Institute of Electrical Engineers of Japan and Wiley Periodicals LLC.