• Novel fusion method combines voltage harmonics using cluster quality weighting. • Perfect topology identification achieved with only 4–6 h of measurement data. • Method remains accurate with 10% measurement error and 60-min time resolution. • High-order harmonics outperform traditional RMS voltage for network identification. • Calinski-Harabasz index weights measurements by their clustering performance. Accurate identification of low-voltage (LV) network topology is becoming increasingly important, as reliable and detailed topological information is vital for effective network operation and precise modelling. Topology identification approaches based on smart-meter data typically rely on RMS voltage, current, and power measurements, which are limited in accuracy due to factors such as time resolution, measurement intervals, and instruments errors. This work introduces a novel methodology for distribution network topology identification through a multi-parametric analysis of smart-meter measurements. The core innovation lies in utilising the Calinski-Harabasz index (CH) as a weighting factor for multi-measurement distance matrices. The proposed framework integrates three distinct classes of measurements: V rms , harmonic components ( V 2 – V 20 ), and THD. The methodology addresses critical challenges in measurement-based topology identification approaches, including high measurement errors, short data collection time intervals, and large time resolution. The resilience of the methodology stems from a hierarchical approach that combines correlation analysis, cluster validation, and graph-theoretic network reconstruction. The results demonstrate significant improvement in the accuracy and robustness of network topology identification, compared to approaches based on single-measurement types.
Othman et al. (Thu,) studied this question.