Abstract: It is very difficult to measure and identify the electromagnetic interference on-site due to the transient signals. Effective identification of the electromagnetic interference for electronic equipment area provides an important basis for the spectral analysis of the radio astronomy site. Base on electromagnetic interference measurement spectrum, an electromagnetic interference (EMI) signal detection and identification method for electronic equipment area on-site is proposed. The multi-group environment spectrum and equipment area spectrum are obtained and preprocessed to realize separation of signal and noise. Additionally, a two-value method is presented to determine the signal boundary and extract interference signal from spectrum. Then, the regression algorithm and the channel occupancy statistical methods are employed to analyze the correlation of signals between equipment area and environment spectrum, then the electromagnetic radiation spectrum of equipment area are identified. The effectiveness and accuracy of this method are verified with tested spectrum coming from Nan Shan 26 m telescope site. More than 94% of the recognition results are in conformity with the artificial recognition results, indicating that this method has high versatility and accuracy.