On-line motor condition monitoring system for abnormality detection

Abstract

An economical and portable on-line motor condition monitoring system based on an advanced microprocessor with a graphic user interface for abnormality detection is described. Both electrical and vibrational signal analysis for fault detection of motors were applied. A Cortex-M4 microcontroller and its built-in high speed analog-to-digital converter were used to process the on-line voltage, current, and vibration signals of the motors during operation. The acquired signals were digitally filtered through an infinite impulse filter and then, fast Fourier transform for spectral analysis was applied for abnormal pattern recognition. The analytic results were displayed in real time on an embedded touch-screen monitor to instantly provide users with the motor's operational conditions. Finally, a prototype was devised and verified through onsite experimentation. Comparing the results to commercial tools showed similar spectral characteristics; the difference for electrical and vibration signals was less than 10%. Thus, the system proposed shows promising ability and feasibility for online use.