Research On Cable Pd Detection And Signal Analysis Based On Improved Built-in Coupling Sensor

With the continuous improvement of the level of urban power grid construction in China,the requirements for the safety and reliability of power grid operation are gradually improved.As the mainstay of urban power grid,XLPE power cable is closely related to the safety and stability of urban power grid operation.As a weak insulation part of the power cable,the cable joint will produce partial discharge due to insulation defects because of its structural complexity and various uncertain factors in the field installation environment.And partial discharge further forms a cable accident,which seriously affects the daily life of residents.At present,in the detection method for partial discharge of a cable,the detection sensor is far away from the local discharge source.This makes the detected signal not only seriously attenuated in amplitude compared to the signal at the local discharge source,but also a certain distortion on the waveform,affecting the accuracy of the signal analysis result.Therefore,in this paper,a detection method of cable partial discharge based on the improved built-in coupling sensor is proposed,which prefabricates the capacitive sensor inside the cable joint to realize the effective detection and identification analysis of the partial discharge signal.The specific research content is as follows:(1)The propagation characteristics of PD signals in high-voltage cables under consideration of the frequency-dependent effects of semi-conducting layers and the skin effect of conductors are studied.Based on the basic theory of electromagnetic field,the propagation mode of electromagnetic wave excited by PD signal in cable is studied,and the influence of incident angle and dielectric constant of material when it propagates at the interface of medium is studied.In the case of considering the frequency-variation effect of the semi-conducting layer and the skin effect of the conductor,the transmission law of the PD signal in the cable is calculated,including the propagation in the cable cross section,the cable body and the cable joint.Which has certain guiding significance for the determination of the installation position and the design of the effective detection frequency band of the sensor.(2)An optimized design method of built-in coupled sensor is proposed and a sensor that meets the requirements of PD detection is developed.The equivalent circuit model of the built-in coupled sensor and the integrated three-dimensional numerical calculation model of the built-in coupling sensor in the cable joint are established.The sensor design scheme is determined by numerical calculation and experimental test results.The built-in coupling sensor was developed and tested for performance.The effective detection frequency band is 1~300MHz,and the detection sensitivity is 5pC.(3)A method for identifying and analyzing partial PD signals under typical defects of cable joints based on vector quantization technique and hidden Markov model is proposed.Four kinds of cable joints are designed and fabricated with typical insulation defects.Based on the built-in coupling sensor,the 110 kV cross-linked polyethylene cable partial discharge test platform is built,and the apparent discharge of different cable joint insulation defects at different voltages is obtained through experiments.Identification and analysis of PD signals under different cable joint defects using vector quantization techniques and Hidden Markov Models that have been successfully applied in the field of speech recognition.And the validity of vector quantization technique and hidden Markov model in cable partial discharge pattern recognition application is verified.