Innovative Approaches to Predicting Surface Subsidence in Inclined Seam Mining: A Comparative Study of Asadi, VNIMI, and KHCNM Functions

Abstract

This study explores the accuracy and applicability of three advanced subsidence prediction methods — Asadi profile function, VNIMI subsidence function, and KHCNM subsidence function — under the challenging conditions of inclined seam mining at Mong Duong coal mine in Vietnam. By comparing predicted subsidence values with direct observation data, the research evaluates the precision of each method using statistical metrics such as actual deviation (Δ), root mean square error (RMSE), mean absolute error (MAE), and correlation coefficient (r). The findings reveal that both the Asadi and VNIMI functions demonstrate high accuracy, with RMSE values of 0.081 m and 0.080 m, and MAE values of 0.055 m and 0.059 m, respectively. The KHCNM function, while slightly less accurate, still provides valuable insights with RMSE and MAE values of 0.101 m and 0.076 m. The study underscores the importance of precise subsidence prediction in safeguarding surface structures and optimizing mining technology, recommending further research to integrate the strengths of Asadi and VNIMI functions for enhanced prediction across different sections of the subsidence basin. This comprehensive evaluation offers a promising pathway for improving subsidence prediction in inclined seam mining, contributing to safer and more efficient mining operations.