基于动态知识规则的多金属矿床储量估算边界智能圈定研究

Focusing on the problem of too much human-computer interaction in ore delineation and orebody blocks division, this study took intelligent ore delineation of polymetallic deposits as example. The optimization solutions to knowledge acquisition, organization and using were systematic explored. The details were as follows: (Ⅰ) Methods of automatic extraction and intelligent transformation for expert knowledge based on characteristics of ore boundaries were discussed, which can dynamic extract expert knowledge in order to improve the accuracy of automatic ore delineation. (Ⅱ) By taking detailed boreholes as the main line, spatial topology as the association, 2D & 3D integrated organization and expression models for mineral resource estimation were constructed, which can enhance the ability of dynamic transfer for orebody section and projection maps. (Ⅲ) Based on the dynamic rules of ore delineation and orebody blocks division, technologies of polymetallic ore delineation were explored. Meanwhile, based on more powers and limited boundaries for polymetallic deposits, the methods of optimal comprehensive blocks division with triangle and polygon were discussed, which can realize intelligent ore delineation and comprehensive blocks division for orebody section and projection maps. This research aims at providing theoretical and technical support for intelligent ore delineation of orebody section and projection maps, which can effectively improve the efficiency and practical level of mineral resource estimation mapping.

针对资源储量估算剖面图、投影图矿体边界圈定及块段划分环节人机知识交互过多的问题，以多金属矿床边界圈定为例，从圈矿知识的获取、组织、运用三个层面提出以下优化方案：(1) 研究基于圈矿特征分析的专家知识自动提取及知识、规则智能转化方法，动态提取较难获取的专家圈矿知识，以提高自动圈矿的准确程度；(2) 搭建以精细钻孔为主线、以空间拓扑为关联，兼顾二维平面与三维空间特性的圈矿数据图库一体化组织模型，以增强剖面图、投影图圈矿数据、知识的动态传递能力；(3) 基于动态获取的圈矿知识规则，研究剖面图、投影图多金属矿体边界智能圈定算法，探讨多权、多限定边界的多金属矿床三角形及多边形矿体块段最优综合划分方法，以实现剖面图、投影图矿体边界的智能圈定及块段自动划分。本课题的研究旨在为矿产资源储量估算边界智能优化圈定奠定理论和技术基础，可有效提升矿产资源储量估算图件的编制效率和实用程度。

矿产资源储量估算边界圈定是资源储量估算图件编制工作中最基本、最重要而又最繁重的核心环节。资源储量估算边界的智能圈定有助于进一步提升资源储量估算图件编制的智能化、自动化水平，对于提高矿产资源勘查精度和效率具有重要作用。.针对资源储量估算剖面图、投影图矿体边界圈定及块段划分的智能优化问题，以多金属矿床边界圈定为例，从圈矿数据、知识的获取、组织、运用三个层面展开优化，研究成果主要包括以下方面：. (1) 通过分析和定量化资源储量估算各阶段圈矿知识规则，研究面向矿体动态构形及矿体形态特征的圈矿知识动态提取方法，构建了面向圈矿对象的增量式圈矿知识规则库，为矿体边界的智能优化圈定提供了依据；. (2) 通过优化剖面图、投影图圈矿数据、知识的组织结构，搭建了以精细钻孔为主线、以空间拓扑为关联的资源储量估算数据二维三维一体化组织与动态表达模型，增强了资源储量估算剖面图、投影图数据的空间动态传递能力；. (3) 围绕资源储量估算剖面图单工程矿体边界圈定和工程间矿体连接两个核心环节展开技术优化，研究了基于动态知识规则的单工程多金属矿体边界智能化动态圈定技术，智能化与专家交互相结合的工程间快速圈矿方法，实现了剖面图多金属矿体边界的智能优化圈定；. (4) 基于GIS、图库一体化、Delaunay三角剖分等技术优化矿体块段划分过程，研究了多权、多限定边界的多金属矿床三角形矿体块段最优划分及储量动态估算新方法，实现了面向矿体空间特征的三角形及多边形矿体块段综合划分、资源储量估算、矿体投影图编制的一体化和自动化。. 本项目研究成果主要运用于资源储量估算、编图软件的研发及相关计算机应用方面，有助于提升资源储量估算及编图的效率、精度和协同化程度，是矿产资源储量估算和编图多维化、动态化、智能化、自动化发展的技术基础，在地矿行业具有广阔的应用前景。