Study On Measurement And Diagnosis Technology Of XLPE Cable Insulation Defects Based On Ultrasonic Method

Cross Linked Polyethylene(XLPE) power cables are widely used in urban power grid due to their advantages of light weight, simply manufacturing process, easy installation, no height restriction, good electrical and thermal properties, and large transmission capacity. However, the operating XLPE power cables will lead to insulation defects and even insulation breakdown accidents caused by the installation process, laying environment, external damage, service condition and other factors. Among of the accidents, the proportion of the insulation failure of the cable intermediate joint and terminal head is more than others. At present, the preventive test is an important means to ensure the reliable operation of the cable. However, the cable insulation is a gradual development process. In the early stage, the discharge signal caused by local defects is very weak and is difficult to be detected by traditional preventive tests. Therefore, the traditional test method has been unable to meet the needs of cable safety operation.Partial Discharge(PD) is the main forms of early insulation faults in XLPE cable accessories, it is not only one of the primary reasons to cause the insulation degradation of cable, but also is the characteristic scalar to describe the insulation state of cable. Therefore, developing the detection method and fault diagnosis of PD in cable accessories have great significance in evaluating the insulation condition immediately, preventing cable insulation failure and ensuring the safety and reliability of the power grid.The detection and diagnosis technology of XLPE power cable insulation based on ultrasonic method was proposed in this paper, and the PD measurement and recognition system was developed. The working principle of this system is: when there was a defect or hidden trouble or even partial discharge under normal working voltage in the terminal or intermediate joint of the operating cables, the electromagnetic wave and ultrasonic signals would be obtained by ultrasonic and ultra high frequency(UHF) combined sensor as PD occurs. The PD signals collected by ultrasonic sensors were amplified and filtered. The characteristics of processed PD signals were extracted by analysis software, the phase-resolved partial discharge(PRPD) spectra were also investigated. At last, the PCA analysis and defect pattern recognition of PD characteristics were carried by Vector Machine Support(SVM) method. Moreover, the location of defect was confirmed by calculating the time delay between the ultrasonic and UHF signals.The detection and diagnosis technology of XLPE cable insulation defects based on ultrasonic method has been studied in this paper, and four results were obtained as follows:(1) The combined sensor which contained ultra high frequency and ultrasonic sensor was developed, then a device of measurement and diagnosis of XLPE cable insulation defects was developed, which contained conbined sensor, data processing unit, and the data acquisition card, analysis and display modules. Experiment showed that by adopting this PD measurement system, the insulation defects were effectively detected; therefore it met the conditions of on-line detection.(2) The phase-resolved partial discharge characteristic parameters extraction method was proposed. The partial discharge spectra of typical defects were obtained and defect recognition fingerprint database was constructed by using this method. Experimental results showed that the proposed method could not only extract the discharge characteristic parameters of each typical defect effectively, but also provide a technical reference for further development of the defect identification.(3) The statistical features ofmax()qH j, and()nH j were investigated. The experimental results indicated that the discharge characteristics of the same defect were similar in the same conditions and had good repeatability. Ultrasonic wave spectra of max()qH j 、 and()nH j corresponding to different types of defects and partial discharge were different, and had obvious distinguishable characteristics. Therefore, these characteristic parameters provide a powerful test basis for a deeper study on the PD mechanism of XPLE cable and the pattern recognition of PD types.(4) Vector Machine Support(SVM) defect recognition method based on principal component feature extraction was proposed. Using this method, the redundant information was removed and the space dimension of the original sample was reduced. Therefore, the accuracy and robustness of the system was improved. The experimental results showed that the typical defects of cable accessory were clearly classified by using this method.The on-line experimental results show that the produced XLPE cable insulation defects detection system has a good performance of providing a test basis for the pattern recognition of PD types and a deeper study on the PD mechanism of XLPE cable.