Gas Insulated Switchgear (GIS) ultra high frequency (UHF) live test presently stands as the most widely employed technique for partial discharge detection in the field. With the ongoing digital transformation of the power grid, the imperative has arisen to advance partial discharge detection toward intelligence and automation. Accordingly, this paper puts forth a digital twin modeling approach for GIS live test, through a comprehensive examination of digital twin technology. This approach encompasses the establishment of a GIS digital twin model, the assessment of both internal and external UHF signals of the GIS, as well as the intelligent guidance for the deployment of UHF sensors, among other aspects. The findings of this study hold significant implications for the maintenance and servicing of GIS equipment.