Research On Fault Diagnosis System For Power Transformer Based On Audio Signal

Transformer is one of the most important equipment in power system. Failure oftransformers may cause great damage to the safety and stability of the system. The faultdiagnose and online monitoring method based on vibro-acoustic signal is of great value tobe exploited since its process is easy to operate and does not affect the normal operation andcauses no harm to the power transformer.This thesis is focused on vibro-acoustic signals with in audio frequency range of20Hz~20000Hz. Common faults of power transformers and their classifications areconcluded according to statistics from China and abroad. The vibration and sound emissionmechanism of transformer core and winding under the normal condition is studied. Thetheoretical results show that the acoustic signal is basically generated bymagnetoconstriction of the core and vibration of winding, which is mainly caused by theelectromagnetic force due to the leakage flux and the current of the coil. Acoustic signalsare collected from transformers in substations, and the regular patterns of transformer soundinduced by theoretical study are certified by the measuring results. Moreover, in order tosimulate insulation faults within transformers in laboratory, corona discharge and sparkdischarge under different conditions are generated, of which acoustic signals are collectedand analyzed with the methods of FFT and STFT, signatures of these signals are gained.According to these studies, a signal acquisition system based on STM32F205RET6isdeveloped, and a software of serial communication and human-computer interaction iswritten in Matlab GUI, making it possible for real-time acquisition of transformer vibrationsignal and processing on PC platform, establishes the foundation for the furtherdevelopment of the fault diagnosis system.