Early Fault Identification of Cable Based on Impulse Voltage

In distribution networks, intermittent early faults often occur before a permanent fault develops in cables. Identifying these early faults is crucial for improving the reliability of power supply in the distribution network. Due to the limitations of neural network-based fault identification methods, this paper proposes an early fault identification technology for cable insulation using impulse voltage. Since impulse voltage carries more energy, it is more likely to reveal hidden defects in cables under normal operating conditions. Firstly, this paper studies the variation patterns of electrical parameters under the early fault state of cables. Subsequently, a cable insulation model and an early fault model are established. Finally, the mechanism of electrical parameter changes when early faults occur in the cable body and its accessories is thoroughly analyzed, leading to the construction of a characteristic impedance spectrum for cables under early fault conditions. The feasibility of the proposed early fault identification technology based on impulse voltage is validated through experimental and simulation results.