The invention discloses a high-voltage circuit interrupter fault diagnosis method based on grey correlation fuzzy clustering. A fuzzy similarity relation matrix is gained by means of a correlation coefficient method, and then a transitive closure operation is conducted. On the basis, clustering analysis is conducted, so that a class which a fault to be diagnosed is in can be gained. By searching for a fault which is similar to the fault to be diagnosed in the class, the component which produces the fault can be gained. The grey correlation analysis method is a powerful tool for resolving a fault diagnosis with little data and weak conditions, and has the advantages of being simple in modeling, little in needed data, and capable of gaining an accurate fault diagnosis under the condition that a confidence level is not very good, thereby providing a basis for reasonable recondition arrangement and safe operation. Large amount of human resource is saved and unnecessary waste is reduced. By adopting the high-voltage circuit interrupter fault diagnosis method based on grey correlation fuzzy clustering in the fault diagnosis of a high-voltage circuit interrupter, the work volume is reduced greatly. The high-voltage circuit interrupter fault diagnosis method based on grey correlation fuzzy clustering has a good application prospect.