Abstract: Driven by the national dual-carbon goals, building a new power system with renewable energy as the main component has become the core task of energy transformation. Relying on the "Green Electricity Town" demonstration project in Minning Town, Yongning County, State Grid Yinchuan Power Supply Company implemented a transparent upgrading and reconstruction program in 68 distribution transformer areas. Following the technical route of panoramic perception and intelligent digital system simulation, a three-level collaborative architecture of "cloud–edge–end" is proposed. It realizes core functions such as virtual measurement of low-voltage branches, precise calculation of milliohm-level impedance, and intelligent outage assessment. A high-precision line loss prediction algorithm based on a bidirectional long short-term memory network is also developed, forming a real-time analysis system covering line loss, reliability, power quality, and load-capacity ratio. Field results show that the abnormal diagnosis accuracy of transformer areas after upgrading exceeds 90%. The proposed approach effectively supports refined line-loss management and provides a replicable "Ningxia model" for building intelligent distribution networks with high renewable energy penetration.