一种地质预测模型构建方法、建模方法、设备及存储介质
By using graph deep learning-based geological prediction models and combining GNN-Transformer models with pre-training strategies, the problem of balancing local and global correlations in 3D geological modeling is solved, enabling the efficient generation of 3D geological models that conform to geological laws.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CHINA UNIV OF GEOSCIENCES (WUHAN)
- Filing Date
- 2022-12-29
- Publication Date
- 2026-07-17
AI Technical Summary
Existing 3D geological modeling methods struggle to balance the local and global spatial correlations of geological elements, and are unable to generate high-quality geological models under the constraints of limited geological data.
A graph deep learning geological prediction model is constructed. By combining the GNN-Transformer model, the spatial relationships of geological nodes are aggregated and graph-structured. Combined with the pre-training strategy, a geological node graph structure mesh is generated, and the model parameters are optimized by the loss function.
It effectively balances the local and global spatial correlations of geological elements, avoids overfitting, generates a three-dimensional geological model that conforms to geological laws, and improves modeling accuracy and efficiency.
Smart Images
Figure CN116245013B_ABST