Analysis and Modelling of Powerline Communication Data for Low Voltage Cable Asset and Condition Management

This paper presents an advanced analytics approach for condition monitoring in low-voltage distribution networks, aimed at detecting degrading and overloaded cables. Attenuation data is collected using a distributed vector network analyzer based on power line communication, enabling continuous, non-intrusive measurements along cable sections. Time-series evaluations are performed using waterfall plots to identify trends and gradual drifts indicative of cable degradation. An HDBSCAN clustering model is then trained to learn the normal operational distribution of each cable section and to detect anomalies. Finally, the paper outlines how these analytical techniques can be integrated in a monitoring system for improved asset management and preventive maintenance.