An Alternative Approach for Condition Monitoring of Brushless DC Motor Drives

In this study, an alternative approach based on redundant and complementary fault signatures is proposed for reliable condition monitoring of a Brushless DC motor drive in critical applications. Focusing on Hall-effect position sensorbased drives, a data-driven approach for the unbalanced system operation diagnosis is presented. To improve the fault detection accuracy under multiple faults, a multi-source database is synthesized, exploiting voltage and current sensors along with piezoelectric sensors for the vibration analysis. The frequency spectra of the electronic controller input and output power signals and the piezoelectric sensor signals are analyzed for the fault diagnosis. Thus, a multi-sensor control scheme is presented and different configurations of misaligned position sensors, with either positive or negative commutation angle errors, are investigated to verify the effectiveness of the proposed fault signatures. Finally, system balancing and commutation signal reconstruction are feasible based on the identification of each position sensor relative and absolute error.