The Statistical Characteristics Of The Measured Intrusive Lightning Waveform In The Station And Its Influence On Air Gap Breakdown Characteristics

Overvoltage characteristics are of great significance to insulation fit,accident tracing,equipment insulation risk assessment and maintenance.At present,the waveform used in the lightning impulse withstands test of in-station equipment is the standard double exponential waveform of 1.2/50μs specified in IEC 60060-1.However,a large number of studies have found that the overvoltage of the lightning intrusion waveform borne by the equipment in the station is significantly different from the standard recommended double exponential waveform,and the insulation resistance characteristics of the equipment under the action of different waveforms are significantly different.In view of this,based on the measured invading lightning waveform in the station,this paper made a detailed statistical analysis of its waveform characteristics,and further explored the influence law of different waveform parameters on air gap breakdown characteristics.The research results in this paper can fill the gap in the study of the overvoltage statistical analysis of the invading lightning waveform in the station,and the insulation characteristics and laws of the air gap under different waveform parameters can provide basic data support for the insulation matching,accident tracing and insulation risk assessment.The main research work and important conclusions are as follows:（1）Based on the measured invading lightning waveform,Fourier transform was used for waveform pretreatment,and the validity of the waveform processing method was proved by the waveform similarity principle.According to the waveform characteristics after preprocessing,three waveform parameters were defined,namely,wave front time,domain oscillation frequency and damping factor,and a peak-zero matching algorithm was proposed to extract waveform parameters,so as to provide effective waveform for subsequent waveform parameter statistical analysis.（2）By testing randomness independence and difference of the waveform samples,the distribution laws of waveform parameters and amplitude characteristics of the overvoltage of 10kV invading lightning waveform was obtained.It was found that the measured invading lightning overvoltage waveform was not the IEC standard 1.2/50μs double exponential waveform,but the attenuated oscillation waveform.The hypothesis test and KS test were used to identify the probability density distribution,and the maximum likelihood method was used to estimate the model parameters.The statistics show that the amplitude distribution of the invading lightning overvoltage of the 10kV side system of the substation was[12,42]kV,and the probability density obeys the Burr distribution.The wave front time distribution was at[8.3,167]μs,the cumulative probability density obeyed Bi Dose Resp distribution,and the oscillation frequency was at[2,22]k Hz.The oscillation frequency of 99.25%was less than 15kHz,and the probability density obeyed the Generalized Parato distribution（-0.18,1.80,4.77）.The damping factor of the waveform was distributed within the range of[0.01,0.90],with a mean value of 0.39,and its probability density obeyed Beta distribution（1.34,2.05）.The amplitude probability density model could be used as the overvoltage probability curve in the statistical insulation matching method,and the statistical distribution results of waveform parameters could provide the waveform and data for the breakdown probability curve test.（3）Based on the statistical analysis results of waveform,the experimental study of air gap breakdown under typical waveform parameters was carried out.Four-factor three-level orthogonal experiments were designed.9 experiments were designed to test the influence of the wave front time,the domain frequency of oscillation and the damping factor on the U_50%breakdown voltage of the air gap by using 5mm needle-plate gap.It was found that the wave front time and the domain oscillation frequency had a significant effect on the breakdown voltage,while the damping factor has no significant effect on the breakdown voltage.Wave front time was the primary factor affecting the air gap U_50%breakdown voltage.With the increase of wave front time,the U_50%breakdown voltage first increases and then decreased.There was a turning point of wave front time,and the turning point of wave front time was related to the domain oscillation frequency.The impact of the domain oscillation frequency was weaker than that of wave front time.With the increase of the domain oscillation frequency,the breakdown voltage first increases and then decreases.There was a turning point of domain oscillation frequency,and the turning point frequency was related to the wave front time.The damping factor had the least influence on the breakdown voltage.As the damping factor increases,the breakdown voltage gradually decreased.When the damping factor exceeds 0.25,the breakdown voltage tended to decrease and become saturated.The support vector machine（SVM）could be used to predict the breakdown voltage of U_50%based on the waveform parameters（wave front time,domain oscillation frequency,damping factor）.The accuracy of the model was 96.71%.