Research On Detection And Diagnosis Of High Voltage Cable Partial Discharge And Dielectric Loss

With the continuous development of China’s power network,the length of high-voltage cables is increasing,and the relationship between the safe and stable operation of high-voltage cables and national production and life is getting closer and closer.Therefore,detection methods that can reflect the running status of the cable in time are needed to reduce the aging of the cable due to insulation defects.The possibility of power accidents ensures the safe and stable operation of the power grid.High-voltage cable partial discharge detection and dielectric loss detection are effective methods in the cable insulation sudden fault detection method.Partial discharge(referred to as PD)monitoring can detect and diagnose local cable faults.Dielectric loss(referred to as dielectric loss)detection can reflect the overall aging degree of the cable.Both of these methods are cable detection methods recommended by the IEEE standard.In this paper,the detection and diagnosis technology of high-voltage cable partial discharge and dielectric loss are studied from four aspects: high-voltage cable compression test experiment,dielectric loss detection technology,partial discharge multi-source identification and dynamic adaptive diagnosis technology.(1)In view of the difficulty in obtaining partial discharge test samples and dielectric loss test sample libraries,a high-voltage cable pressurization test experimental platform based on typical defects is established,On this test platform,the cable partial discharge and dielectric loss detection test are carried out.In this paper,by selecting and setting five typical defect types,the high-voltage cable single-source defect and multi-source defect squeeze test experiments are carried out to obtain PD data.In this paper,the test of dielectric loss of high-voltage cable is carried out to obtain the data of dielectric loss detection.The starting voltage and extinction voltage in the partial discharge detection of high-voltage cables based on single-source defects and multi-source defects are analyzed in the paper.Then,the shape and amplitude law of partial discharge pulses caused by single-source defects and multi-source defects at different voltage levels are analyzed.(2)In view of the difficulty of high-voltage cable dielectric loss live detection,a multipoint synchronization-based high-voltage cable dielectric loss detection technology and accuracy improvement method is proposed in this paper.The high-voltage cable dielectric loss detection methods are summarized in this paper,which divides the high-voltage cable dielectric loss method into absolute measurement method and relative measurement method.The causes of the high-voltage cable dielectric loss detection error are analyzed in the paper.Then,an improved method accordingly is proposed.A new type of high-voltage cable dielectric loss detection system is presented,and the effects of leakage current separation,reference voltage compensation,and the installation of a temperature measurement system on the error of the dielectric loss detection value are analyzed.(3)Aiming at the difficulty of local multi-source identification of high-voltage cables,a multi-source identification method for high-voltage cables based on high-reliability synchronous trigger acquisition and feature optimization selection is proposed.In this paper,a multi-source partial discharge method based on high-reliability synchronous trigger acquisition and multi-source information fusion is proposed for multi-source partial discharge generated at different positions of the same cable.In this paper,for the multi-source partial discharge caused by different defects in the same position of the cable,a multi-source partial discharge identification method based on feature optimization selection and clustering segmentation is proposed.In this paper,a multi-source partial discharge method for high-voltage cable based on optimized features and convolutional neural network deep learning is proposed for multisource partial discharge generated by cables and power equipment connected by cables.(4)Aiming at the difficulty of diagnosing the state of high-voltage cables based on partial discharge detection,a dynamic adaptive state diagnosis technology for high-voltage cables based on partial discharge detection technology is proposed.This paper analyzes the influencing factors of high-voltage cable status diagnosis,and divides the influencing factors of highvoltage cable status diagnosis into three types: considering the cable’s own information,the identification of PD interference,and the source and type of PD.In this paper,combined with the influencing factors of high-voltage cable condition diagnosis,a high-voltage cable condition diagnosis model based on dynamic adaptive technology is proposed.Five high-voltage cables of a power supply company,and that it is necessary to pay attention to the cable are diagnosed.However,due to the problem of power outage maintenance schedule,the cable has an accident.In this paper,flashover occurred during the diagnosis of an EPR cable.By analyzing the changes in the amplitude of partial discharge before and after flashover and the number of partial discharge pulses,new influencing factors for the diagnosis of high-voltage cable status were obtained.