Scholarlink

<p>The sedimentary environments, lithological frameworks, and geological properties of the coal seams in the North China Coalfield exhibit regional heterogeneity. Therefore, conventional unified formulas often fail to capture the regional nuances required for predicting coal seam floor failure depth. This study categorizes the North China coalfield into three representative geological blocks and used the Bayesian analysis, ordered regression analysis, and nonlinear regression analysis models to predict floor failure depth, based on six key influential factors. The improved ordered regression formula reduced the average prediction error by 5.72 compared to linear regression, the improved Bayesian formula by 0.08 compared to standard empirical formulas, and the nonlinear regression formula by 0.96 compared to multi-layer perceptron predictions. The obtained prediction formulas for different blocks exhibited superior regional applicability and accuracy, offering a practical approach for accurately predicting coal seam floor failure depth in North China and providing valuable insights and references for predicting failure depth of coal seam floor in similar geological environments.</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

Multi-model Zonal Prediction of Floor Failure Depth in the North China Coalfield

  • 期刊论文

  • 发表日期：

    2026年4月20日

查看全文
  • Zhaojun Song,
  • Jinjin Tian,
  • Jiawen Wang,
  • Xulai Jin,
  • Dongjing Xu,
  • Wenming Liu

摘要

The sedimentary environments, lithological frameworks, and geological properties of the coal seams in the North China Coalfield exhibit regional heterogeneity. Therefore, conventional unified formulas often fail to capture the regional nuances required for predicting coal seam floor failure depth. This study categorizes the North China coalfield into three representative geological blocks and used the Bayesian analysis, ordered regression analysis, and nonlinear regression analysis models to predict floor failure depth, based on six key influential factors. The improved ordered regression formula reduced the average prediction error by 5.72 compared to linear regression, the improved Bayesian formula by 0.08 compared to standard empirical formulas, and the nonlinear regression formula by 0.96 compared to multi-layer perceptron predictions. The obtained prediction formulas for different blocks exhibited superior regional applicability and accuracy, offering a practical approach for accurately predicting coal seam floor failure depth in North China and providing valuable insights and references for predicting failure depth of coal seam floor in similar geological environments.