Multiple fault diagnosis in variable speed drives through current measurements

In this paper, a signal-based method is proposed for the detection and isolation of open-switch faults in variable speed drives. The fault diagnosis algorithm relies only on the information of stator currents. Residuals and features are generated from the trajectory of Park's current vector and symmetrical components of line currents. Fault detection is achieved by directly monitoring the magnitude of Park's vector. Meanwhile, an isolation algorithm based on pattern recognition is proposed by applying a supervise classification technique. The proposed scheme is able to evaluate 18 single and simultaneous faults conditions under steady-state operation. Experimental data is evaluated from a test bench of 1 HP induction motor to validate the fault detection and isolation scheme presented in this work.