Research On Winding Fault Detection Of Electromagnetic Voltage Transformer Excitated By High Voltage Pulse

Electromagnetic voltage transformer is an important equipment in power system,which can detect voltage,measure power and provide relay protection signal.It has the advantages of simple structure,low cost,small error,large transmission capacity and good transient response,so it is widely used in high voltage distribution network and power plant.However,there may be defects such as bubbles and insulation damage on the surface of enameled wire in the manufacturing process of electromagnetic voltage transformer,or the winding may suffer from overvoltage impact during operation,which will lead to the decrease of winding insulation.It may cause faults such as inter-turn short circuit,inter-layer short circuit or inter-layer discharge in the winding,resulting in inaccurate measurement of the transformer and affecting the safe and stable operation of the power system.The existing detection methods mostly use the correlation coefficient of waveform data for fault judgment,but this method requires more experimental samples,and the discrimination of slight winding faults needs to be improved.Therefore,studying a method for detecting and analyzing the winding faults of electromagnetic voltage transformers with higher discrimination makes the detection results more accurate,which is of great significance for reliable power supply in power systems.Firstly,the structural characteristics of electromagnetic voltage transformer were expounded,and the equivalent circuit model of high voltage winding was established.In addition,the voltage distribution in winding under high voltage pulse excitation was analyzed.The fault detection scheme was given by using the change of the equivalent capacitance when the winding fault occured.The winding was subjected to an impulse test,and the discrete wavelet transform was used to denoise and high-frequency analyze the response voltage.The wavelet energy entropy of the impulse response voltage waveform was obtained,which was used as a characteristic parameter for winding fault diagnosis.Secondly,in order to verify the effectiveness of wavelet energy entropy as a characteristic parameter for insulation fault judgment,an all-solid-state Marx high-voltage pulse generator was designed to generate high-voltage pulses with little interference and standard lightning waveforms for electromagnetic voltage transformer winding fault detection.The main circuit,synchronous trigger circuit,drive circuit and protection circuit of the generator were designed.The main circuit and auxiliary circuit were simulated by PSpice simulation software,and the rationality of the circuit structure and component parameters was verified.Then the pulse generator was fabricated and tested,and the output 1.0～1.5 k V standard lightning waveform high voltage pulse met the design requirements.Finally,an experimental model of electromagnetic voltage transformer was made,and the fault simulation of the winding was carried out.The inter turn short circuit,inter layer short circuit,and inter layer discharge faults were set by taking out the winding taps.Using the specially developed all-solid-state Marx high-voltage pulse generator,impact experiments were conducted on the experimental model of the transformer.The differences and characteristics of the impulse response voltage waveforms between the normal winding and the fault winding were analyzed.The fault diagnosis of the winding was performed using wavelet energy entropy,and the diagnostic effect was compared with that of the correlation coefficient.The results show that the wavelet energy entropy has obvious discrimination of various winding faults under different voltages,and the detection effect is better than the correlation coefficient,which can better judge the winding insulation state.