Scholarlink

<p>Small-scale faults associated with major faults in geological structure are difficult to measure accurately by existing geological exploration techniques. Identifying key parameters, such as fault strike and shape, is critical for optimizing mining operations. This study introduces a novel method for the precise detection of small-scale faults within coal seams based on data from coalbed gas drainage boreholes. The borehole data were preprocessed by considering the elevation phase difference of the faults and the similarity in prior coal burial depths. The particle swarm optimization (PSO) algorithm enhanced the nearest neighbor propagation clustering, achieving 80.56% accuracy, while the k-means and k-medoid algorithms reached 94.44% and 100%, respectively, on the same borehole data. Clustering results from the PSO-k-means algorithm (PKMA) were used to fit anomalies, with deviations between the fitted and actual dips being 3.92°, 3.12°, and 2.54° for Faults 1, 2, and 3, respectively. The PKMA method excelled in identifying 1-m drop faults as borehole spacing increased to 8 m and 10 m. ​​Field validation through borehole data from the Huainan Coal field confirmed the reliability of PKMA, showing strong agreement between predicted fault geometries and on-site geological exposures.​​ These results further validate the feasibility of the proposed small-scale fault identification method based on drill hole data. This study used geologic information embedded in little-attended coalbed gas drainage boreholes to provide new insight for precise measure small-scale faults in underground coal mines.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Application and Validation of PSO-k-Means Clustering for Small-Scale Fault Identification in Coal Mines: A Case Study of the Huainan Coalfield

  • 期刊论文

  • 发表日期：

    2025年10月15日

查看全文
  • Baocai Wang,
  • Chunhui Cheng,
  • Qianting Hu,
  • Yongjiang Luo,
  • Xinshang Hou,
  • Rongxing Zhang

摘要

Small-scale faults associated with major faults in geological structure are difficult to measure accurately by existing geological exploration techniques. Identifying key parameters, such as fault strike and shape, is critical for optimizing mining operations. This study introduces a novel method for the precise detection of small-scale faults within coal seams based on data from coalbed gas drainage boreholes. The borehole data were preprocessed by considering the elevation phase difference of the faults and the similarity in prior coal burial depths. The particle swarm optimization (PSO) algorithm enhanced the nearest neighbor propagation clustering, achieving 80.56% accuracy, while the k-means and k-medoid algorithms reached 94.44% and 100%, respectively, on the same borehole data. Clustering results from the PSO-k-means algorithm (PKMA) were used to fit anomalies, with deviations between the fitted and actual dips being 3.92°, 3.12°, and 2.54° for Faults 1, 2, and 3, respectively. The PKMA method excelled in identifying 1-m drop faults as borehole spacing increased to 8 m and 10 m. ​​Field validation through borehole data from the Huainan Coal field confirmed the reliability of PKMA, showing strong agreement between predicted fault geometries and on-site geological exposures.​​ These results further validate the feasibility of the proposed small-scale fault identification method based on drill hole data. This study used geologic information embedded in little-attended coalbed gas drainage boreholes to provide new insight for precise measure small-scale faults in underground coal mines.