Evaluation Of Cable Joint Internal Faults Based On Impedance Information In The Frequency Domain

In the modern power system,the cable assumes the responsibility for the transmission of strong electrical energy,and plays an extremely important role.With the rise in the number of cable lines,the accidents of power cable happened frequently.The accidents occurred more often in cable joint,which causes serious economic losses and adverse social impact.The method of evaluating for the internal defects of cable joints is particularly important,which will help to prevent the permanent fault of the cable.Impedance,as the property of the cable itself,is related only to its own characteristics,such as the shape and geometry of the cable,the relative position of the conductor and the dielectric material.It can be seen that the electrical impedance of the defective cable will be different compare to the normal one.From the basic principle of transmission line,it is easy to get the conclusion that when electromagnetic wave propagates along the line,the refraction and reflection are inevitable at the node,where the line parameters have changed.A novel detection method has been proposed,in which the impedance information of the cable port collected at high frequency is employed to achieve the diagnosis of cable defects.Unfortunately,when this method is applied to the cable joint,the calculation of impedance can be very complicated due to the complex structure of cable joints,and there is still a lack of an effective method to evaluate the internal defects of the cable joint through the frequency domain information of impedance.In view of the shortcoming of the existing research,a method for evaluating the internal defects of cable joints based on impedance frequency-domain information was proposed.The main research and innovation were as follows:(1)A calculation model of cable joint distribution parameter was proposed.For the complex structure of the cable joint,this paper focuses on the equivalent and separation of the cable joint model,meanwhile,the skin effect of conductors and the dielectric properties of semi-conductive materials at high-frequency are considered.Through then establishment of a 3D finite element-based simulation model,the accuracy of this calculation model was verified.(2)Based on the impedance in frequency-domain and the integral transformation function,a diagnostic function was proposed which can availably characterize the internal defect degree of the cable joint.According to the analysis of the cable joint diagnosis function with different defect levels and different locations,the similarities and differences between the two diagnosis functions are compared,which provides a basis for the subsequent fault classification.Finally,a confirmatory experiment was taken,which built with the vector network analyzer and a 10 kV cable,to verify the method proposed.(3)A evaluation model of cable joint internal defect was proposed,which based on the support vector machine.This evaluation model is based on the rule that the wave of diagnostic function changes with the degree of internal defect,meanwhile the extremum and the ripple amplitude are used as input information.In order to improve the accuracy,the particle swarm optimization algorithm was employed to optimize the parameters and the optimal kernel function is selected.