Optical fiber sensing and partial discharge fault diagnosis for power cable health monitoring

Abstract Partial discharge (PD) in power cables is a significant precursor to cable insulation aging and failure. Accurately monitoring PD is crucial for ensuring the safe and stable operation of power systems. This paper proposes a method for detecting PD in power cables based on a fiber-optic Michelson interferometer. This method leverages the high sensitivity and electromagnetic interference resistance of the fiber-optic Michelson interferometer to achieve real-time detection of acoustic signals generated by PD in cables. To address the noise interference present during the detection process, this paper designs an effective noise reduction algorithm to improve the signal-to-noise ratio. Subsequently, time–frequency (TF) analysis techniques are used to extract features from the de-noised signal. The resulting TF images are then used as a training dataset for a deep learning model, enabling automatic recognition and classification of PD signals. Experimental results demonstrate that the proposed method can effectively improve the accuracy of PD detection, validating the potential application of a fiber-optic Michelson interferometer combined with signal processing and deep learning techniques in monitoring PD in power cables.