| In view of the excellent insulation performance,XLPE cables are widely used in power systems.However,limited to insufficiency of manufacturing process and harsh running environment,insulation defects bring great hidden dangers to the security,stabilization,and reliability of the power system.The detection of cable insulation status has become an important research field in power system.In the current detection method,withstand voltage test is destructive,AC superposition method is susceptible to interference,and the partial discharge signal is difficult to recognize.In order to accurately detect the insulation performance,a fully-covered interdigital capacitive sensing technology is proposed,which can be used for long-term nondestructive monitoring of cable insulation,especially for the cable joints.Design and parameter optimization of the sensor is studied in 2D and 3D model.First of all,the simulation model of the sensor is built with the Comsol Multiphysics software by finite element.The sensor has the advantages of more concentrated internal electric field,less external interference,and larger measurement area.Three performance indicators,i.e.signal strength,sensitivity,and penetration depth,are defined for sensor optimization.Secondly,the impact of single parameter changes on the performance indicators are analyzed separately in the 2D and 3D model.Then,through full factorial design,the effect degree of parameter changes on the indicators is comprehensively researched to determine the design criteria.Finally,the reasons for the difference between 3D and 2D simulation data are investigated,and electrode parameters are optimized to reduce the difference.Simulation shows that compared to traditional sensing technology,the fully-covered capacitive sensor is more suitable for long-term monitoring.The design criteria for the sensor are:firstly,the penetration depth is determined according to the insulation thickness to determine the number of the electrode pair,and secondly,considering the sensitivity,the larger electrode width-gap ratio K is selected.Then,based on the signal strength,the thinner PI is selected as the substrate material.The sensor electrode thickness D1 can be 18-36μm.The electrode length is 5 cm,after that the sensor sensitivity no longer changes with the length.The axial edge electric field is the reason for the difference between the 2D and 3D data.The formula is used to explain the change of sensitivity with length.The sensors are used to detect the insulation defects.Firstly,the sensors with different electrode parameters are made by PCB technology,and an experimental platform is built to determine the experimental sweep frequency.Secondly,by setting the reference frequency,the difference between the simulated value and the experimental value is compared and analyzed.Finally,the defect models are made for the experiment,according to the damage of the insulation layer and the jacket layer and the dislocation defect of the stress cone at the cable joint.The greater the number of electrode pairs,the greater the signal strength,the smaller the sensitivity and penetration depth;the greater the electrode gap ratio,the greater the signal strength and sensitivity,the smaller the penetration depth.The experimental results show that 2-200kHz is the stable frequency to measure the insulation capacitance.When the electrode width gap ratio is equal to 1,the best electrode parameter can effectively weaken the difference of experiment and simulation caused by the gap between electrode and cable.The capacitance value of the defects is smaller than the normal value.For the defects in insulation and jacket,the change in capacitance caused by the damage of the jacket layer is greater than that of the insulation layer;The change in capacitance caused by the dislocation defect of the stress cone is larger,so the capacitive sensing technology can recognize the difference between the dislocation defect and the jacket and insulation damage.For different cable defects,the variation of cable insulation capacitance is not different with the increase of frequency. |