A method and system for predicting the production and transportation capacity of a geological gas storage by fusing image data
By employing image text recognition technology and a multi-level adaptive ensemble random forest model, the problem of insufficient data utilization in the prediction of geological gas storage capacity was solved, achieving high-precision and high-stability prediction results.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- HUAZHONG UNIV OF SCI & TECH
- Filing Date
- 2026-05-08
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies for predicting the production and transportation capacity of geological gas storage facilities rely on a single data acquisition channel, cannot effectively utilize unstructured image data, and have incomplete feature engineering, resulting in insufficient prediction accuracy and stability, making it difficult to meet the demand for efficient and accurate prediction.
Image text recognition technology is used to analyze unstructured image data, and a multi-level feature engineering and multi-level adaptive ensemble random forest model is constructed. Combined with dual-engine hyperparameter optimization, multi-source data and dynamic environmental features are integrated to construct a multi-level adaptive sampling rate ensemble random forest model (MSR-RF) to achieve accurate prediction of the gas extraction and transportation capacity of geological gas storage facilities.
It achieves prediction stability in low-noise environments and generalization potential in high-noise environments, improving the prediction accuracy and stability of geological gas storage capacity and meeting the needs for efficient and accurate prediction.
Smart Images
Figure CN122416136A_ABST