A rapid moisture detection system and accurate calculation method suitable for coal gangue
By deeply integrating contact near-field dielectric spectroscopy measurement with non-contact GNSS-R remote sensing measurement, and combining deep residual networks and physical information neural networks, the accuracy and speed issues of coal gangue moisture detection have been solved, achieving rapid and accurate moisture detection.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- HUACHAOHONG (GUANGZHOU) NEW ENERGY TECHNOLOGY CO LTD
- Filing Date
- 2026-04-23
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies for detecting moisture in coal gangue suffer from problems such as inaccurate detection, slow speed, and inability to achieve continuous monitoring. In particular, the complex composition and variable physical state of coal gangue lead to poor generalization ability of single detection principles and traditional models.
A deep fusion of contact near-field dielectric spectrum measurement and non-contact GNSS-R remote sensing measurement was adopted, combined with a dual-channel deep residual network and a physical information neural network. An adaptive weighted fusion of near-field dielectric constant spectrum data and GNSS-R signal characteristic parameters was performed to construct a coal gangue moisture inversion model. A mineral dielectric prior knowledge base and Maxwell-Garnett effective medium theory were introduced as physical constraints.
It enables rapid and accurate detection of moisture in coal gangue, effectively resists interference caused by density and composition fluctuations, improves detection accuracy and model generalization ability, and provides confidence indexes for self-evaluation of detection results.
Smart Images
Figure CN122409770A_ABST