On-line fault detection of inter-turn short circuits in turbo-generator rotor windings

The on-line fault diagnosis technique based on search coil method for detecting inter-turn short circuits in turbo-generator rotor winding is used. Based on completely different properties of induced electric potential signals in the air gap and noise characteristic of the generator operated in different conditions is analyzed, wavelet transform and Mallatpsilas local maximum de-noising algorithm is presented, thereby realizes the detection of shorted turns fault and finds the fault slot position in turbo-generator rotor windings. Theoretical analysis and experimental results show that the method is particularly suitable for on-line fault detection of inter-turn short circuit in turbo-generator rotor windings.