<p>For ventilation calculations in underground mines, mine roadway dimensions are important. The key dimensions are cross-sectional areas and perimeters. Three-dimensional scanning technology allows for acquiring a point cloud that reflects the roadway outline. However, this technology not allow for direct calculation of average values of cross-sectional area and perimeter along the length of the roadway. The study is based on filtered point cloud data and excludes the effects of roadway support structures and similar complex factors. It focuses on extracting geometric features and developing a segmentation method. The proposed method uses multi-parameter fusion of three-dimensional point clouds to segment roadway features, allowing accurate fitting to regular shapes. Seven roadway features are integrated using principal component analysis. Kernel density estimation then establishes thresholds for roadway segmentation. This makes it possible to divide the roadway into segments that have similar morphology and unique contour shapes. The proposed contour shapes effectively represent the roadway contours instead of the point clouds. Parameters such as <InlineEquation ID="IEq1"> <EquationSource Format="TEX">$\overline{{O }_{{d}_{s}}}$</EquationSource> <EquationSource Format="MATHML"><math> <mover> <msub> <mi>O</mi> <msub> <mi>d</mi> <mi>s</mi> </msub> </msub> <mo>¯</mo> </mover> </math></EquationSource> </InlineEquation>, <InlineEquation ID="IEq2"> <EquationSource Format="TEX">$\overline{RC{L }_{s}}$</EquationSource> <EquationSource Format="MATHML"><math> <mover> <mrow> <mi>R</mi> <mi>C</mi> <msub> <mi>L</mi> <mi>s</mi> </msub> </mrow> <mo>¯</mo> </mover> </math></EquationSource> </InlineEquation>, and <InlineEquation ID="IEq3"> <EquationSource Format="TEX">$\overline{{V }_{os}}$</EquationSource> <EquationSource Format="MATHML"><math> <mover> <msub> <mi>V</mi> <mrow> <mi mathvariant="italic">os</mi> </mrow> </msub> <mo>¯</mo> </mover> </math></EquationSource> </InlineEquation> are introduced to compare the accuracy of the contours determined by segmentation versus those determined for the non-segmented roadway. The method was experimentally verified to improve the accuracy of airflow resistance calculation when using the segmented contours. The example of a scanned roadway is used to verify the segmentation method. The results show that the contours separated by this method fit the point clouds better and more accurately represent the geometric features of the roadway. This method is significant for evaluating roadway geometric morphology for intelligent air resistance calculations in mine ventilation systems.</p>

Method for Roadway Cross-Section Determination and Airflow Resistance Calculation Based on Multidimensional Feature Fusion of Point Cloud Data

期刊论文
出版模式：

开放获取
发表日期：

2025年10月6日

查看全文

Zhipeng Qi,
Dariusz Obracaj,
Kazimierz Piergies,
Marek Korzec,
Ke Gao

摘要

For ventilation calculations in underground mines, mine roadway dimensions are important. The key dimensions are cross-sectional areas and perimeters. Three-dimensional scanning technology allows for acquiring a point cloud that reflects the roadway outline. However, this technology not allow for direct calculation of average values of cross-sectional area and perimeter along the length of the roadway. The study is based on filtered point cloud data and excludes the effects of roadway support structures and similar complex factors. It focuses on extracting geometric features and developing a segmentation method. The proposed method uses multi-parameter fusion of three-dimensional point clouds to segment roadway features, allowing accurate fitting to regular shapes. Seven roadway features are integrated using principal component analysis. Kernel density estimation then establishes thresholds for roadway segmentation. This makes it possible to divide the roadway into segments that have similar morphology and unique contour shapes. The proposed contour shapes effectively represent the roadway contours instead of the point clouds. Parameters such as $\overline{{O }_{{d}_{s}}}$ $\bar{O_{d_{s}}}$ , $\overline{RC{L }_{s}}$ $\bar{R C L_{s}}$ , and $\overline{{V }_{os}}$ $\bar{V_{os}}$ are introduced to compare the accuracy of the contours determined by segmentation versus those determined for the non-segmented roadway. The method was experimentally verified to improve the accuracy of airflow resistance calculation when using the segmented contours. The example of a scanned roadway is used to verify the segmentation method. The results show that the contours separated by this method fit the point clouds better and more accurately represent the geometric features of the roadway. This method is significant for evaluating roadway geometric morphology for intelligent air resistance calculations in mine ventilation systems.