Inversion of In-situ stress field in near-fault coal mining area by coupling numerical simulation with deep learning

Understanding the in-situ stress distribution is crucial for coal mine production. In mining area 105 of Yuandian No. 1 Coal Mine, in-situ stress measurements were carried out using the stress relief method. Based on the collected data, the in-situ stress field was inverted using a Transformer archi...

Full description

Saved in:
Bibliographic Details
Main Authors: Rongqin Deng, Yuan Zhang, Dong Wu, Jiakun Lv, Yanan Jing, Zheng Zhen, Peng Shi
Format: Article
Language:English
Published: Taylor & Francis Group 2025-12-01
Series:Geomatics, Natural Hazards & Risk
Subjects:
Online Access:https://www.tandfonline.com/doi/10.1080/19475705.2025.2545374
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Understanding the in-situ stress distribution is crucial for coal mine production. In mining area 105 of Yuandian No. 1 Coal Mine, in-situ stress measurements were carried out using the stress relief method. Based on the collected data, the in-situ stress field was inverted using a Transformer architecture. The inversion results indicate that the Transformer achieves higher accuracy than the PSO-SVR method, demonstrating its suitability for in-situ stress inversion in near-fault mining areas. The distribution and characteristics of in-situ stress are jointly influenced by the Wugouyangliu-1 fault and burial depth; however, at greater depths, burial depth becomes the primary controlling factor. A significant stress concentration is observed near the fault zone, with stress values increasing by approximately 60%. Moreover, the axes of mining roadways are almost perpendicular to the maximum horizontal principal stress, which adversely affects roadway stability. This work verifies the feasibility of Transformer for in-situ stress field inversion in the near-fault coal mining area.
ISSN:1947-5705
1947-5713