The insulation performance of high-voltage equipment is one of the important indicators for its safe and reliable operation. Ultra high frequency (UHF) detection method has become one of the detection methods of Partial discharge (PD) due to its high sensitivity and strong anti-interference ability, but it is powerless to deal with periodic narrow-band interference in the UHF frequency band. Aiming at the problem of mixed periodic narrow-band interference in UHF partial discharge signals, it is proposed to use Fast Fourier transform (FFT) to integrate empirical mode decomposition, ensemble empirical mode decomposition (EEMD) and wavelet threshold noise reduction method combined anti-jamming algorithm. Finally, the algorithm is verified through experiments, and compared with the anti-jamming results of EEMD, wavelet hard threshold, soft threshold and pauta criterion, it is found that the wavelet soft threshold has the best anti-jamming effect, which proves that the algorithm has a good anti-jamming effect in UHFPD detection.