Cable Aging Type Identification Method Based on Frequency Domain Reflection and Multilayer Perceptron

In order to solve the problem that current cable inspection methods are difficult to recognize the type of defects in cables, this paper provides a method for identifying the aging types of cable based on the Frequency Domain Reflection (FDR) and Multilayer Perceptron (MLP). Firstly, sample cables are made for heating and electrical aging of the cables. Using FDR measures the reflection coefficient of the cable head before and after aging. Secondly, the effective parameters in the measurement data are analyzed and extracted to establish the original sample library of cable aging faults. Finally, the MLP is introduced to realize the fast and accurate judgment of cable aging faults. To verify the effectiveness of the method, tests are conducted on sample cables. The test results show that the cable status can be accurately assessed using this evaluation model, with an average accuracy rate of about 95 %. The method provides a non-invasive, high-precision solution for cable defect type identification.