Cable fault detection based on enhanced spread spectrum time domain reflectometry

Spread spectrum time domain reflectometry (SSTDR) is applicable for cable fault detection, effectively ensuring the safety of power generation systems. However, the superposition effect between the reflected and incident signals can significantly disrupt the correlation calculation between these signals. To address this issue, this paper introduces a novel cable fault detection method based on enhanced spread spectrum time domain reflectometry (ESSTDR). This approach removes the incident signal component from the acquired waveform, thereby mitigating its interference with the reflected signal and enhancing reflective characteristics. Additionally, a relative difference amplitude index is proposed to quantitatively assess the noise suppression capability of this method. Through comparative experiments encompassing various fault levels, this study demonstrates that ESSTDR outperforms SSTDR in terms of detection performance. This paper also proposes a fault severity assessment model based on the double exponential function and compares the performance of different fault features. The results show that the standard deviation of the reflected waveform is more suitable for reflecting the fault severity.