Abstract
Power Line Communication (PLC) harnesses the existing infrastructure of power lines for data transmission. As one application, PLC is being used for monitoring and control in distribution networks. In this paper, we propose an autonomous technique that exploits the communication channel estimated inside legacy PLC modems to determine the health of distribution cables. In particular, we consider paper insulated lead covered (PILC) cables widely used in low and medium voltage distribution networks that are most susceptible to thermal degradations. Measurement campaigns have shown that these thermal degradations cause dielectric property changes in PILC cable insulations, which also result in changes in PLC channel conditions. However, through channel characterization of healthy and degraded cables, we demonstrate that the estimated channel frequency responses are not sufficiently distinctive for manual diagnosis. We therefore propose a machine-learning based technique that not only achieves our set target, but is also able to estimate the cable health under varying load conditions. Simulation results show that our proposed technique accurately estimates thermal degradation severities in PILC cables. We thus believe that PLC based cable health monitoring can be used as an autonomous remote diagnostics method that can be integrated into a smart-grid concept and has the promise of being more cost-effective than deploying personnel and/or dedicated equipment.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call