Goaf Fine Identification and Spatial Form Reconstruction Method Based on 3D Seismic Attribute Fusion

In order to improve the goaf identification accuracy and form restoration ability under complex geological conditions, a type of identification modeling method based on 3D seismic multi-attribute fusion is proposed. By conducting optimization selection on sensitive attributes such as amplitude, frequency, phase and others, a comprehensive response index (CI) is constructed, combined with K-means clustering and discrete smooth interpolation, goaf boundary extraction and 3D modeling are achieved. Compared with the single amplitude method, the boundary error of CI method is reduced by 48%; Compared to the PCA+SVM fusion method, the validation matching degree is improved by 7%. The actual measurement results indicate that the amplitude attenuation in the goaf reaches 38%, the frequency decreases by 11 Hz, the spatial positioning error is less than 2 m, and the matching degree exceeds 93%. This method significantly enhances the accuracy and modeling integrality of goaf identification, providing technical support for the safe mining in mining areas.