In view of the lack of a solution to deal with abrupt-changing fault and abnormal disturbance signals in the existing transformer fault diagnosis methods, this paper presents a collaborative diagnosis of abrupt-changing fault of electronic instrument transformer based on wavelet transform and WVD. Based on the idea of station-area, the instrument transformers in substations are divided into cooperative discriminate groups, and the principle of wavelet transform modulus maxima is used to locate the sudden change time of signals. In the case of multiple transformers disturbing signal due to system disturbance, the fault transformer can still be identified by the good time-frequency distribution characteristics of WVD distribution. A simulation model is built in PSCAD to verify the effectiveness of the proposed method in identifying fault transformers when abrupt-changing fault and system disturbance occur simultaneously or at different times. This work is supported by National Natural Science Foundation of China (No. 51207050).