Research On Cable Defect Location Based On Reflection Coefficient Spectrum

Power cables play an irreplaceable mission for power transmission in large-scale power systems.In the actual operation process,the cable will accelerate the degradation of the performance of the cable insulation layer due to defects such as local aging or damage,and even cause insulation failure.The main reasons for local defects of cables are: due to local temperature rise,local damage and moisture,partial discharge to form electrical breakdown,deformation and damage caused by insulation layer stretching,scratching,extrusion or foreign object penetration,and even nuclear damage.Radiation causes local aging of the insulating layer.At this stage,the raw data of cables to be collected for cable defect location is complicated,and the required frequency is too high,which is not conducive to on-site testing.To this end,this paper combines the frequency domain reflectometry to carry out research on the proposed problem.In order to solve the shortcomings of too much raw data and too high frequency of the existing cable local defect localization method,a local defect localization method based on the cable reflection coefficient spectrum selection kernel function for integral transformation is proposed.Firstly,the calculation model of the reflection coefficient spectrum at the head end of the cable with local defects is established by using the distribution parameters of the cable.After that,the influence mechanism of the local fault on the reflection coefficient spectrum of the cable is explained,and the integral transformation of the reflection coefficient spectrum is further studied.The principle of cable positioning.In order to reduce the required raw cable data and sampling requirements,a cable fault location method is proposed by integrating and transforming the reflection coefficient spectrum.Then,the simulation positioning analysis is carried out for cables with different defect conditions,and the influence of different signal frequency ranges and the number of sampling points in the frequency domain on the positioning results is obtained.The results show that the method can still achieve high-precision positioning while reducing the signal frequency and the number of sampling points,and the error is less than 1%.Finally,by reducing the sampling quantity of the original data in the actual experiment,the effectiveness of the method in field application is demonstrated,and it has high practical engineering value.In view of the above problems,this paper proposes a method for diagnosing local defects of 10 kV cables in distribution network by using the generalized orthogonal integral based on the reflection coefficient spectrum.The reliability of the proposed method has the advantages of less required raw data,low frequency range and fewer sampling points,and has certain advantages in practical engineering applications.Therefore,the cable defect diagnosis method based on reflection coefficient spectrum proposed in this paper has certain value.