Abstract: To solve the problem that the partial discharge (PD) signals of high voltage power equipment is seriously affected by the periodic narrowband interferences, a narrowband interference suppression of the PD signals based on the generalized S transform and the mathematical morphology is proposed in this paper. The proposed method first separates the time-frequency spectrum images of the original PD signals and the narrowband interferences from the noise-polluted PD signals, accurately identifying the original PD signals and the narrowband interferences in the time-frequency domain and avoiding the problem that the strong narrowband interferences cover the time-frequency characteristics of the original PD signals. Then, the total least squares-estimation of the signal parameters via the rotational invariance technique is used to accurately estimate the characteristic parameters of narrowband interferences in order to suppress the narrowband interferences in the noise-polluted PD signals. The narrowband interference suppression tests are performed in the simulated and the measured PD signals. From the test results, the proposed method succeeds in effectively separating the narrowband interference signals and the PD signals from the time-frequency spectrum images of the noise-polluted PD signals, improving the accuracy of time-frequency analysis. Compared with the traditional FFT spectrum minimum entropy deconvolution filtering method and the adaptive singular value decomposition method, the proposed method effectively suppresses the narrowband interferences with small amplitudes, obtaining a better narrowband interference suppression result and restoring the waveform characteristics of the PD signals.