Research On Defect Modeling And Intelligent Identification Of Urban Underground Transmission Cable Joints

High voltage DC cables are one of the main means of transporting electricity in urban underground cables.110 kV cross-linked polyethylene(XLPE)cables are commonly used in urban underground cables,and have become the main form of high voltage DC cables due to their excellent electrical and mechanical characteristics.In recent years,cable production technology has developed rapidly,but research into insulation degradation and failure problems that occur when cables are in service is now lagging behind,often increasing the difficulties of power maintenance.Due to the cable manufacturing process,the cable cross-section is subject to metal residues,semiconductor residues,moisture and other faults in the cable joint connection,which can lead to partial discharges(PD),an important manifestation of cable faults and insulation deterioration,and can reflect the operating condition of the cable.In order to detect cable operating problems in time and ensure the safety of the electrical system,this article examines cable faults and partial discharges in cables.(1)This thesis discusses cable construction,types of cable defects,and types of partial discharges.The causes of partial discharges are discussed and the mechanism of partial discharges is analysed in detail.The main factors influencing partial discharges in cables are investigated as electric field strength,space charge density,and partial discharge mapping.(2)Based on the finite element simulation software to simulate the cable defects,simulation of its different defects under the electric field strength,space charge density,the simulation results of the analysis,the analysis shows that the metal defects by the electric field strength and space charge density more influence.(3)In this thesis,a multi-scale neural network incorporating a Convolutional Block Attention Module(CBAM)is proposed for cable defect recognition in order to better detect the ground insulation status of cables during operation and classify and identify both electric field strength and space charge density data,based on two types of The data are analysed collaboratively to achieve good recognition results,with an overall recognition accuracy of 96.59%.(4)Partial discharge mapping is used to classify and identify the forms of partial discharges caused by cable defects,based on the partial discharge mapping obtained from online monitoring of cable operation.The partial discharge mapping includes Phase Resolved Partial Discharge(PRPD)and Phase Resolved Pluse Sequence(PRPS),from which the characteristics of the mapping are extracted to distinguish the types of partial discharges and to identify partial discharge phenomena.To ensure the safe operation of underground cables in cities.