Summary

Generator Circuit Breakers (GCB) is a critical component in large-scale hydropower and nuclear power stations, characterized by high rated currents and the capacity to break shortcircuit currents with high DC components. During interruption, high-energy arcs cause severe erosion to moving contacts, significantly affecting the equipment's electrical life. Since contacts are sealed within arc chambers, direct observation is impossible. Consequently, engineering relies on indirect evaluation methods like loop resistance measurement, which tend to be inaccurate. To address this, this paper proposes a novel indirect assessment method based on

"vibration fingerprints". The friction generated between contacts during operation produces distinct vibration signals. These signals evolve as electro-erosion alters the surface morphology.

By establishing a mapping relationship between feature evolution and erosion degree, the extent of damage can be inferred.

This study utilized a laboratory electrical life test of a GCB prototype and analyzed vibration signals from closing operations due to their superior stability. Drawing on Recurrence

Quantification Analysis (RQA), a Vibration Fingerprint RQA (VFRQA) model was developed to extract texture features from Recurrence Plots (RPs). Results indicate that metrics specifically Recurrence Rate (RR), Longest Diagonal Line Length (Lmax), and Longest Vertical

Line Length (Vmax) exhibit a significant correlation with contact erosion severity. Furthermore, standardized Euclidean distance was employed to quantify signal deviation and establish early warning thresholds. In conclusion, VFRQA provides a non-intrusive reference approach for

GCB electro-erosion assessment, with its feasibility preliminarily verified by the test results of this study. Future research will focus on addressing sensor interference and integrating artificial intelligence with multi-dimensional data including breaking current and material properties to construct a comprehensive electro-erosion assessment model.

Additional informations

Publication type Session Materials
Reference D1_12524_2026
Publication year
Publisher CIGRE
Country China, People's Republic of
Study committees
File size 2 MB
Price for non member 30 €
Price for member 30 €

Authors

JIANG Zongmin - Xi’an High Voltage Apparatus Research Institute Co., Ltd.; LI Yuanchao - Xi’an High Voltage Apparatus Research Institute Co., Ltd.

Keywords

Generator Circuit Breaker, Contact, Vibration Fingerprints, RQA, Electro-erosion Assessment

Contact Electro-erosion Assessment of Generator Circuit Breaker Based on Vibration-fingerprint-RQA