钻孔插值建模方法因流程自动化、模型更新快速化的特点在地下工程三维建模场景中得到广泛应用,而较为稀疏、数据有缺失的原始钻孔数据限制了建模的准确性,且其他非钻孔类的地质数据难以有效利用。针对以上问题,提出一种基于迭代插值和约束修正的虚拟钻孔自动化构建方法,同时考虑数据空间分布均匀性、非钻孔数据有效约束及插值数据的复利用,以获得更高质量的虚拟钻孔数据,提高建模准确性。首先,提出针对空间虚拟点位均匀获取的三角化加密方法,获得钻孔加密系数μ;其次,提出搜索和迭代插值算法实现虚拟钻孔分层信息获取,并引入多源数据约束修正准则和异常值剔除修正准则对迭代算法进行优化;同时,提出了基于加权平均表征的模型准确性评估方法。基于营盘壕煤矿延安组地层数据,采用提出的方法进行了建模验证,结果表明:“迭代插值”和“约束修正”分别使模型的层厚误差、层厚相对误差、埋深误差减少“3.12 m、1.39%、1.09 m”和“0.42 m、0.19%、0.41 m”。
The borehole interpolation modeling method is widely used in 3D modeling of underground space engineering scenarios due to its characteristics of process automation and rapid model updates. However, the sparse and missing original borehole data limit the accuracy of modeling, and other geological data is difficult to effectively utilize. A virtual borehole automation construction method based on iterative interpolation and constraint correction is proposed to address the above issues, which considers the uniformity of data space distribution, effective constraints of non-borehole data, and the compounding of interpolation data, to obtain higher quality virtual borehole data and improve modeling accuracy. Firstly, a triangulation encryption method is proposed for obtaining uniformly distributed virtual borehole positions, which is characterized by borehole encryption coefficients; Secondly, a search and iterative interpolation algorithm is proposed to obtain virtual borehole layering information, and a multi-source data constraint correction criterion and outlier removal correction criterion are introduced to optimize the iterative algorithm; Meanwhile, a model accuracy evaluation method based on weighted average representation was proposed. Based on the geological data of the Yan'an Formation in Yingpanhao Coal Mine, the proposed method was used for modeling verification. The results showed that “iterative interpolation” and “constraint correction” reduced the layer thickness error, layer thickness relative error, and burial depth error of the model by “3.12 m, 1.39%, 1.09 m” and “0.42 m, 0.19%, 0.41 m”, respectively.
[1] 李璐,刘新根,吴蔚博.基于钻孔数据的三维地层建模关键技术[J]. 岩土力学,2018,39(3):1056-1062.
[2] 杜子纯,刘镇,明伟华等.城市级三维地质建模的统一地层序列方法[J]. 岩土力学,2019,40(增1):259-266.
[3] 肖建勇,严伟,乔世范,等.基于BIM的公路边坡可视化管理方法研究[J]. 铁道科学与工程学报,2024,21(6):2342-2358.
[4] 李健,刘沛溶,梁转信等.多源数据融合的规则体元分裂三维地质建模方法[J]. 岩土力学,2021,42(4):1170-1177.
[5] 雷传扬,刘兆鑫,文辉,等.基于多源数据和先验知识约束的复杂地质体三维建模研究[J]. 地质论评,2022,68(4):1393-1411.
[6] 董前林,李青元,曹代勇,等.多源地质数据综合三维建模及地质学分析——以青海木里三露天勘探区为例[J]. 煤田地质与勘探,2017,45(3):37-44.
[7] 郭甲腾,刘寅贺,韩英夫,等.基于机器学习的钻孔数据隐式三维地质建模方法[J]. 东北大学学报(自然科学版),2019,40(9):1337-1342.
[8] 熊玖琦,刘星.基于支持向量机的顾北煤矿北一矿区三维地质模型构建[J]. 科学技术与工程,2022,22(19):8194-8199.
[9] 徐学闯,张恒兵,符韶华,等.基于钻孔伪标签半监督深度学习的沈阳市浅表地层三维建模研究[J]. 地理与地理信息科学,2023,39(6):9-17.
[10] 张士军.钻探岩心缺失原因及可能解决方法的分析[J]. 吉林地质,2011,30(2):131-133.
[11] 宋仁波,崔璐阳,陶叶青, 等.基于GIS空间分析的虚拟钻孔确定与实现方法[J]. 煤田地质与勘探,2018,46(6):67-73.
[12] 王波,雷传扬,刘兆鑫,等.三维地质建模过程中综合地质剖面构建方法研究[J]. 沉积与特提斯地质,2021,41(1):112-120.
[13] 张瀚,桂蕾,王腾飞,等.基于BP神经网络的第四系覆盖物厚度预测及三维地质建模[J]. 地球科学,2024,49(2):550-559.
[14] 薛涛,包训栓,朱小弟,等.多源数据三维地质结构模型约束的属性建模方法:以北京市通州城市副中心为例[J]. 地学前缘,2023,30(3):529-536.
[15] 孙黎明,刘禹杉,张睿卓,等.基于几何细分的三维地质模型自适应精细化构建方法[J]. 岩土工程学报,2023,45(增1):244-248.
[16] Geuzaine C, Remacle J F. Gmsh: A3-D finite element mesh generator with built-in pre-and post-processing facilities[J]. International Journal for Numerical Methods in Engineering,2009,79(11): 1309-1331.
