Abstract: To meet the demands of intelligent operation and maintenance as well as unmanned transformation in new energy photovoltaic power stations, this paper designs and implements a multi-source data fusion analysis platform for unmanned booster stations. The platform adopts an overall architecture of "cloud-edge collaboration and hierarchical decoupling", integrating comprehensive automation, photovoltaic arrays, energy storage, and other dispersed heterogeneous systems within the booster station through standardized data access protocols. Based on a unified data middle platform and AI middle platform, the platform establishes core functional modules such as panoramic monitoring, equipment state prediction, operational optimization, and intelligent inspection, achieving deep data integration and intelligent applications. Engineering applications demonstrate that the platform effectively addresses information silos, significantly improves operation and maintenance efficiency and power station economic performance, providing a reliable technical solution and practical pathway to support the new operation and maintenance model of "unmanned and remote centralized control".