Reducing the Impact of False Alarms in Induction Motor Fault Diagnosis

摘要: Early detection and diagnosis of incipient faults is desirable for on-line condition assessment, product quality assurance, improved operational efficiency induction motors. At the same time, reducing probability false alarms increases confidence equipment owners in this new technology. In paper a model-based fault system recently proposed by authors motors experimentally compared alarm performance with more traditional signal-based motor estimator. addition to nameplate information required initial set-up, uses measured terminal currents voltages, speed. The model embedded empirically obtained using dynamic recurrent neural networks, resulting residuals are processed wavelet packet decomposition. effectiveness detecting most widely encountered electrical mechanical faults, while minimizing impact from power supply load variations, demonstrated through extensive testing staged faults. scalable different ratings it has been successfully tested data 2.2 kW. 373 kW, 597 kW