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