Abstract: To achieve precise monitoring of overvoltage, this study designs an intelligent transient overvoltage detection method for a UHV gas-insulated substation. The research introduces Mayer and Cassie models to establish an arc model for the substation. Sensor technology is applied for voltage signal acquisition, while signal feature extraction is conducted based on the characteristics of ultra-high voltage gas-insulated substations. Comparative experimental results demonstrate that the proposed method achieves high-precision transient overvoltage monitoring with excellent consistency between theoretical predictions and actual measurements.