Abstract: From the current situation, the medium-voltage distribution grid has established a relatively complete reliability management system, while the low-voltage distribution grid is still in the pilot exploration stage, mainly facing three issues: difficulty in topological connectivity, poor data quality, and insufficient depth of intelligent application. In response to the above issues, this paper proposes a technical route of "medium-voltage empowerment, low-voltage extension, event connectivity, and intelligent drive", and constructs a corresponding technical system focusing on four aspects: dynamic topology automatic identification, low voltage full-link perception, multi-source data fusion and event collaborative aggregation, and intelligent algorithm application. A power-industry large model is further introduced at the intelligent algorithm layer to support regulation knowledge retrieval, fault semantic understanding, disposal plan generation, and human-machine collaborative verification. The research results have been verified on a large scale in Heilongjiang, Jiangsu, and other places. According to the verification results, the accuracy rate of topology identification has reached 100%, the linkage time of medium and low-voltage events has been shortened from hours to seconds, and the line-loss rate has also significantly decreased.