Optimization Study of Multidimensional Computational Model for Distribution Network Cable Fault Location and Hidden Danger Warning System in Smart Grid Environment

Due to the limited gain-bandwidth product of traditional spectrum analyzers, the output voltage at high frequencies is too low, resulting in significant signal attenuation when testing long-distance cables. This leads to distortion in the impedance spectrum at high frequencies. This paper proposes a cable fault diagnosis and localization algorithm based on a pseudo-trapezoidal wave excitation impedance spectrum digital reconstruction method. It analyzes the impact of different fault types on the local fault impedance of power cables, the attenuation characteristics of the excitation voltage signal at the cable input, and the power transmission characteristics at different voltage amplitudes. A high-capacity pseudo-trapezoidal excitation system based on a SiC high-speed inverter is designed to measure impedance spectrum information up to a maximum frequency of 7 MHz, utilizing the impedance spectrum digital reconstruction method to achieve full coverage across a wide frequency range from 0.1 Hz to 60 MHz. Compared to traditional low-voltage sine wave excitation, the localization results using this space-time transformation function exhibit smaller interval oscillations and faster convergence. The fault localization accuracy improves by 80%, validating the effectiveness and accuracy of the proposed method.