JIANG Hui, LI Kui, GUO Qingbin, et al. Performance Degradation Evaluation Method of Low-voltage DC Circuit Breaker Based on Morphological Characteristics of Gas-generating Components[J]. 2026, 46(10): 4325-4336.
JIANG Hui, LI Kui, GUO Qingbin, et al. Performance Degradation Evaluation Method of Low-voltage DC Circuit Breaker Based on Morphological Characteristics of Gas-generating Components[J]. 2026, 46(10): 4325-4336.DOI: 10.13334/j.0258-8013.pcsee.250532.
Low-voltage DC circuit breakers are important for power systems
and their performance degradation may lead to system faults. As the core component of the arc extinguishing system
the degradation of gas-generating materials is closely related to the performance of circuit breakers. A performance degradation evaluation method for low-voltage DC circuit breakers based on the morphological characteristics of the gas-generating components is proposed. First
the working principle and morphological changes of the gas-generating component are analyzed. The variation characteristics of the arc erosion marks on the back of the gas-generating components and the edge ablation area on their side are found
both of which are used as key indicators to evaluate the degradation of circuit breakers. The K-means method with fixed clustering centers is applied for the image segmentation of the arc erosion marks. A quantitative characterization method for the severity of arc erosion marks is proposed to reveal the variation characteristics of arc erosion marks under different breaking times. The edge contour line of the side image of the gas-generating component is extracted to analyze the arc ablation area. The quantitative characterization method of the side arc burning degree is constructed
and the variation characteristics of the side arc burning area under different interruption times are revealed. By integrating the back and side degradation features of the gas-generating components
a grading method for the performance degradation degree of the low-voltage DC circuit breaker is proposed
addressing the problem of performance degradation evaluation during service. Experimental validation shows that the prediction accuracy reaches 94%.