Digital limestone mine intelligent management system
By combining multimodal perception and data acquisition, basic physical model construction, physical information fusion and solution engine, model topology self-consistent evolution and intelligent decision-making and causal auditing modules, a high-fidelity digital twin of the limestone mine management system was realized. This solved the problem of the lack of physical mechanism explanation and quantitative causal attribution in the prediction model in the existing technology, and realized the dynamic evolution of the mine status and quantitative causal attribution.
CN122365249APending Publication Date: 2026-07-10YANZHOU SINOMA CONSTR CO LTD
View PDF 0 Cites 0 Cited by
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- YANZHOU SINOMA CONSTR CO LTD
- Filing Date
- 2026-04-10
- Publication Date
- 2026-07-10
Smart Images

Figure CN122365249A_ABST
Abstract
The application relates to the technical field of mine safety monitoring and data processing, and discloses a digital limestone mine intelligent management system, which comprises a multi-modal perception and data acquisition module, a basic physical model construction module, a physical information fusion and solving engine and a model topology self-consistent evolution module. The multi-modal perception and data acquisition module is used for acquiring real-time observation data of a mine. The basic physical model construction module is used for defining a partial differential equation system of the mine. The physical information fusion and solving engine is used for fusing data and equations through a physical information neural network, generating a physical state field and a residual field. The model topology self-consistent evolution module is used for identifying an incoherent area according to the residual field and correcting the equation system. The intelligent decision and causal audit module is used for carrying out prospective deduction and counterfactual audit based on the physical state field. The application integrates a physical model defined by a partial differential equation and a physical information neural network, and through a composite loss function, the system cooperatively optimizes neural network parameters to simultaneously fit real observation data and follow physical laws, thereby generating a state field which is continuous in a global domain and physically self-consistent.
Need to check novelty before this filing date? Find Prior Art