Self-developed "Dimeng" Engine Powers Trial Operation of Daya Bay Nuclear Power's Real-Scene 3D Twin Platform

Recently, the real-scene 3D twin platform for Daya Bay Nuclear Power, jointly developed by Daya Bay Nuclear Power and Zhike/Simulation Company under CGN Digital Technology, was launched for trial operation. Leveraging the self-developed "Dimeng" domestic 3D engine from the Simulation Company, the platform integrates multi-source data including 3D models, Gaussian splatting, forward design data, and GIS for the Daya Bay Nuclear Power base.

The "Dimeng" engine is a graphics development tool independently developed by the Simulation Company to address the characteristics of the nuclear power industry, such as dense equipment, complex system interconnections, and scattered data, providing underlying capabilities including model rendering, spatial interaction, physical simulation, and data-driven processing. Building on this foundation, the Simulation Company, in collaboration with Daya Bay Nuclear Power, constructed the real-scene 3D twin platform. Based on the engine, a three-tier data architecture of "plant level—building level—equipment level" was established, integrating data interfaces across design, operation and maintenance, and management phases, enabling real-time mapping of site-wide data and coordinated control between the physical plant and the virtual plant. The platform supports functions such as "cloud inspection" of buildings and one-click equipment positioning. With hotspot tags, operation and maintenance personnel can complete equipment positioning and access key data within 10 seconds, whereas traditional methods require cross-system, step-by-step retrieval averaging approximately 3 minutes. With the platform, the workload for a single equipment information query is significantly reduced, greatly improving efficiency.

This platform marks the first full-site deployment of digital twin technology in the Group's nuclear power operation and maintenance sector. Going forward, the Simulation Company will expand application scenarios based on this platform, promoting the scaling of digital twins from a single-base pilot to large-scale replication, and supporting the Group's digital transformation. The trial operation of this platform signifies a solid step forward in empowering "virtual-physical symbiosis" intelligent operation and maintenance with a self-developed 3D engine. In the later phase, the project team will continue to deepen digital twin applications, comprehensively promote existing experience to multiple units, and add new features such as video fusion, system twinning, hazardous chemical management, and 3D pipeline visualization, building a solid foundation platform for digital transformation.