Research on isomorphic fusion of multi-source geological data for open-pit coal mines

Transparent geological conditions are essential for intelligent open-pit coal mining. To address insufficient exploration accuracy and weak integration of heterogeneous geological data, this study proposes a multi-source geological data isomorphism method for open-pit coal mines. The method analyzes the spatial variability of coal-rock strata and geological structures, evaluates the applicability and complementarity of different datasets, and applies a stepwise pairwise fusion strategy: Kriging fusion of geological maps and boreholes, Zoeppritz-equation-based fusion of boreholes and seismic data, and Faust-equation-based fusion of seismic and electromagnetic data. The fused outputs were then comprehensively integrated to construct a stratigraphically and structurally screened dataset. Taking Dananhu No. 2 Open-Pit Coal Mine as a case, data for coal seams, other strata, Quaternary deposits, faults, burnt rock, and groundwater were processed. Case-specific evaluation used a locked external holdout dataset and four benchmark methods. To avoid overgeneralizing single-case numerical outcomes, performance was expressed using four ordinal grades (Excellent, Good, Moderate, and Poor) across five accuracy dimensions and six engineering-efficiency/cost dimensions. The proposed method was graded Excellent across all five accuracy dimensions; for engineering performance, it was graded Moderate in preliminary data-collation and single-run computational efficiency and Excellent in repeated-correction efficiency, overall engineering efficiency, cost control, and accuracy-cost suitability. The resulting model maintained interpreted structural and stratigraphic discontinuities while improving continuity where supported by observations; surface smoothness was treated only as a continuity diagnostic. These grades are case-specific and require site-specific recalibration and multi-mine validation before broader generalization.