The invention provides an electronic transformer abnormality identification method. The method includes the following steps that: dividing cooperative identification transformer groups; inquiring a cooperative identification group where a target transformer is located; establishing an electronic transformer output signal mathematical model; performing wavelet singular value detection; and performing WVD time-frequency analysis. According to the electronic transformer abnormality identification method provided by the technical schemes provided by the present invention, the information sharing characteristic of an intelligent substation is fully utilized, data redundancy is improved, and the reliability of the algorithm is further increased.