Title:  Stray Flux Monitoring and Multi-Sensor Fusion Condition Monitoring for Squirrel Cage Induction Machines

Committee:

Dr. Thomas Habetler, ECE, Chair, Advisor

Dr. Lukas Graber, ECE

Dr. Daniel Molzahn, ECE

Dr. Bo He, Ansys

Dr. Rhett Mayor, ME

Abstract: This research work investigates the ability of external magnetic flux-based condition monitoring to detect rotor-related and incipient stage bearing faults in squirrelcage induction machines (SCIMs). This work also discusses the multisensory synergy of the external magnetic flux measurement with other measurements. To investigate the stray flux-based monitoring technique, this dissertation presents a theoretical analysis of the characteristic components in the stray flux spectrum of SCIMs as well as experimental validations. A sensor fusion method to efficiently utilize the information from heterogeneous sensor measurements (external magnetic flux and stator current) to achieve higher rotor-related fault detection sensitivity and a higher fault type recognition rate is presented. A wavelet packet decomposition adaptive threshold denoising method is proposed for flux-based incipient bearing fault detection. A novel sensor fusion-based rotor vibration observer method is proposed. The observer can reject the electrical disturbances from the supply side. Meanwhile, the proposed observer is less affected by the mechanical noise from lousy environment than using vibration-based monitoring.