The Application Research And The Mode Recognition Technique Of The Electrical Power Cable

With the increasingly widespread application of XLPE power cables, the occurrence of insulation failureis constantly accompanied. During the process of failuresettlement, the insulationdetective technology has made a big progress.In this paper, a combination of theoreticalmethods and experimental methods,based on the existing XLPE partial discharge theory and the insulationdetective technology, has been usedto study the partial discharge pattern recognition method for XLPE power cables.Firstly, pattern recognition system has been described in detail in this paper,including the basic principles of partial discharge pattern recognition.For the BPNN neural network using PRPD mode statistical operator and the SART neural network using PRPD modedischarge times and statistical operator, this paper designed two artificial neural network pattern recognition program.By designing artificial neural network to identify the measurement signal, we receive the following results: The recognition rate of BPNN neural network usingnumber of discharges of PRPD mode as its input reaches 88%, the main factoris the occurrence of partial discharge strength. The recognition rate of the BPNN neural network usingstatistical operator of the PRPD model as input reaches95%. The recognition rate of the SART neural network using discharge times and statistical operator as inputreaches 98 %, the main factoris the sample quantities. In this paper, the 110 KV cable failure analysis of Benxi Pingshan line demonstrate the effectiveness of partial dischargepattern recognition method.