Detection and Identification of Cable Aging Using Power Line Communication Technology

Nondestructive inspection methods that use broadband reflecting signals to identify microscopic flaws and aging characteristics in power cables are often employed, however, they suffer from limited detection distances and a single detection target. This study proposes a scheme for harnessing power line communication (PLC) devices to accomplish the detection and identification of multiple cable aging. A model of the propagation of a high-frequency carrier signal through a cable is first combined with mathematical models of various types of aging, such as water-tree aging, electric-tree aging, and thermal aging, to simulate the network channel state in the process of different insulation degradation of the cable. The channel frequency domain response at the receiver can then reflect the incidence of cable aging, and the proposed approach, in conjunction with machine learning (ML) algorithms, can recognize different types and severity of aging under interference. Computer simulations and field experiments were conducted to validate the viability of the suggested approach, and the results shown that 98% correctness could be attained in on-line monitoring despite interference considerations. In theory, the range is doubled when compared to the reflected signal approach with the same bandwidth and power.