Partial Discharge Pattern Recognition Of XLPE Based On Fractal Characteristic

Currently, cross linked pollythene power calbe is used more and more extensively on net, its voltage grade is much higher and that its breakdown is also increase .On the basis of the method of on-line PD monitoring and XLPE fault detection and theory connecting with experiment ,this paper studies gray image of PD signal created by XLPE cable ,calculates fractal characteristic number and recognize the pattern of PD signal with ANN.This paper introduces the condition of XLPE cable development, analyses the reason of PD signal created by XLPE power cable ,collects and analyses the method of XLPE power cable's fault detection ,and then explains the utility meaning of researching XLPE cable's PD.This paper analyses the character of PD impulse created by XLPE cable- magnitude is only μA, time of duration is very short, its spectrum is between 10kHz and100MHz.Large energy is put to study a method , which is used a current coupling sensor of high frequency and broadband to detect the PD . The experiment proves that the band with of this sensor can be 39kHz and 43MHz and can detect the PD signal of XLPE cable effectively.Aimed at insulation defects which appear in real operation of XLPE power cable, this paper produces various insulation defects and do the power aged experiment of four defects under power voltage. PD goes along with the time of put on voltage and the times and modulus of discharge.This paper studies fractal theory and estimation method of specific dimension, from the fractal character of PD. The pattern recognition system of PD gray image recognizing characters get better results .The practicability that fractal theory is used in PD recognition of XLPE cable is verified.