Abstract <p>The detection and location of partial discharge are crucial for assessing the insulation condition of equipment efficiently, which is of great significance to the safety and stability operation of electrical equipment. However, accurate time delay estimation is the prerequisite for high precision localization of partial discharge. To address the challenge of degraded delay estimation accuracy under low signal-to-noise ratio conditions, this paper proposes a novel generalized cross-correlation algorithm incorporating PHAT-SCOT joint weighting. A double exponential oscillation attenuation model is employed to simulate characteristic partial discharge pulse signals. The original signal and the delay signal are firstly cross-correlated and the power spectral density function is obtained by PHAT-SCOT joint weighting. Then, the inverse Fourier transform is carried out and the delay estimate value is obtained by peak detection. The proposed algorithm is comparatively evaluated against conventional cross-correlation and quadratic cross-correlation methods through statistical analysis of delay estimation accuracy rates and mean square error metrics. An experimental system of partial discharge signal detection is implemented to validate the proposed methodology through practical time delay estimation experiments. The simulation and experimental results show that the algorithm can sharpen the peak value of cross-correlation function under the condition of low SNR and that it has the advantages of high accuracy of delay estimation and small root mean square error, so it can be applied to partial discharge detection and location.</p>