Line impedance anomaly detection based on OPTICS

The accurate identification of distribution substation area topology and precise diagnosis of potential hazards in low-voltage distribution networks are essential for advancing intelligent power distribution and enhancing grid security. Traditional approaches based on fixed threshold rules for voltage feature analysis and impedance anomaly detection often struggle with limited adaptability to complex and dynamic data environments, resulting in constrained recognition accuracy and applicability. To address these challenges, this paper proposes a topology identification method based on kernel K-means clustering and an impedance anomaly detection approach utilizing the OPTICS (Ordering Points To Identify the Clustering Structure) algorithm. The proposed approach first performs spatial clustering by exploiting the inherent similarity in voltage measurements among smart electricity meters, thereby overcoming the rigidness of threshold-based methods. Topological structures are then identified through impedance analysis. In parallel, the OPTICS algorithm resolves the issues of missed and false judgments inherent in traditional threshold approaches, ensuring superior detection accuracy. Experimental evaluations and on-site validations confirm that the proposed framework eliminates the reliance on manual set thresholds for topology recognition in low-voltage distribution networks and enables high-precision detection of impedance anomalies. This establishes a reliable and application-oriented technical foundation for early fault warning and improved operational insight in distribution systems.