Scholarlink

<p>Many orebodies in hydrothermal deposits exhibit complex morphologies, such as network and lens-shaped geometries, that are closely associated with ore-controlling structures. Conventional 3D implicit modeling often leads to anomalous extrapolation of orebody boundaries under conditions of sparse sampling and insufficient mineralization constraints. This paper proposes an implicit modeling framework that (1) evaluates model extrapolation error (<i>U</i>) by integrating ore-related information and (2) adaptively supplements off-surface constraints for the model’s iterative extrapolation to construct the high-quality 3D orebody model. To reliably quantify <i>U</i> and identify low-confidence model zones, we employed an improved anisotropic inverse distance weighting (AIDW) that the mineralization constraints (e.g., fault features) are embedded into AIDW parameter estimation, with anisotropy optimized via particle swarm optimization. Off-surface points are selected as supplementary constraints based on the <i>U</i> data, and non-zero implicit function values are assigned to these points within the framework of generalized radial basis function. The density and the constraint strength coefficient of off-surface points are adjusted progressively to dynamically balance local details and global structure, limiting unreasonable shape extensions in anomalous extrapolated areas. The proposed method was validated in the Canzhuang orogenic gold deposit, Eastern China, which includes 2818 drillholes, with model accuracy evaluated based on orebody affiliation probability error (<i>U</i> value), area-to-volume (SA:V) ratio, and agreement with explicit model. The results showed that, compared to traditional methods of Hermite radial basis function, the proposed method reduces extrapolation errors (<i>U</i> value) by approximately 72% and improves model quality (SA:V ratio) by about 67%. The improved model better aligns with explicit modeling in the sparsely drilled regions, preserving complex geometric features such as vein- and lens-shaped structures, while effectively suppressing issues such as model abrupt truncation and inflation. Therefore, the proposed 3D implicit modeling is a promising tool for 3D modeling of various hydrothermal deposits.</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

Iterative 3D Implicit Orebody Modeling Based on Automatic Refinement of Off-Surface Points

  • 期刊论文

  • 发表日期：

    2026年7月2日

查看全文
  • Xiancheng Mao,
  • Yuanqian Nie,
  • Zhankun Liu,
  • Hao Deng,
  • Jin Chen,
  • Xiaowen Han

摘要

Many orebodies in hydrothermal deposits exhibit complex morphologies, such as network and lens-shaped geometries, that are closely associated with ore-controlling structures. Conventional 3D implicit modeling often leads to anomalous extrapolation of orebody boundaries under conditions of sparse sampling and insufficient mineralization constraints. This paper proposes an implicit modeling framework that (1) evaluates model extrapolation error (U) by integrating ore-related information and (2) adaptively supplements off-surface constraints for the model’s iterative extrapolation to construct the high-quality 3D orebody model. To reliably quantify U and identify low-confidence model zones, we employed an improved anisotropic inverse distance weighting (AIDW) that the mineralization constraints (e.g., fault features) are embedded into AIDW parameter estimation, with anisotropy optimized via particle swarm optimization. Off-surface points are selected as supplementary constraints based on the U data, and non-zero implicit function values are assigned to these points within the framework of generalized radial basis function. The density and the constraint strength coefficient of off-surface points are adjusted progressively to dynamically balance local details and global structure, limiting unreasonable shape extensions in anomalous extrapolated areas. The proposed method was validated in the Canzhuang orogenic gold deposit, Eastern China, which includes 2818 drillholes, with model accuracy evaluated based on orebody affiliation probability error (U value), area-to-volume (SA:V) ratio, and agreement with explicit model. The results showed that, compared to traditional methods of Hermite radial basis function, the proposed method reduces extrapolation errors (U value) by approximately 72% and improves model quality (SA:V ratio) by about 67%. The improved model better aligns with explicit modeling in the sparsely drilled regions, preserving complex geometric features such as vein- and lens-shaped structures, while effectively suppressing issues such as model abrupt truncation and inflation. Therefore, the proposed 3D implicit modeling is a promising tool for 3D modeling of various hydrothermal deposits.