Detection of AC machines insulation health state based on evaluation of switching transients using two current sensors and eigenanalysis-based parameter estimation

Application of monitoring systems is growing in modern drive applications to ensure continuous operability. Insulation health state can be examined offline by evaluating the transient reaction to a voltage step excitation. Using the inverter as a source of excitation, it is possible to perform an insulation test by evaluating the resulting transient current sensor signals available in modern drives. The trace of the machine´s transient current reaction is influenced by parameters like the parasitic winding capacitances. Due to degradation of the winding insulation these parasitic parameters change and as a consequence also the shape of the transient current signal is changed. As the degradation induced changes in the signal are no very prominent, measurement procedure and signal processing play an important role in the estimation of the insulation health state. In the following, a method is presented to derive an insulation health state indicator based on the availability of two phase current sensors only and using eigenanalysis of the transient signal frequencies. The performance of the health state indicator derived is verified on a 1.4MW induction machine with preformed stator coils equipped with additional tapings.