Abstract: This study addresses the issue of abnormal parameter identification and diagnosis in preventive tests of high-voltage circuit breakers by proposing a comprehensive approach that integrates multidimensional feature analysis with intelligent diagnostic models. On this basis, a knowledge-driven diagnostic model based on expert systems and a data-driven classification model based on support vector machines are constructed. The case analysis demonstrates that the proposed method can effectively identify early faults that are difficult to detect with traditional threshold methods, achieve precise fault location through the collaborative analysis of multi-dimensional data, and provide reliable technical support for the condition assessment of high-voltage circuit breakers.