Intelligent prediction system for surface movement and deformation in the subsequent filling mining of inclined, thick, and large ore

Abstract

Abstract Guizhou is a mountainous region that faces numerous engineering challenges during the mining of thick inclined ore bodies, including complex geological conditions, limited monitoring systems, and insufficient on-site monitoring. Developing intelligent methods for predicting surface movement and deformation in mining areas under such geological conditions is critical for preventing ecological damage, safeguarding lives and property, and ensuring safe mining operations. This study focuses on the subsequent filling mining of inclined thick ore bodies as the engineering context. Using the probability integral method, an integrated system was developed in MATLAB App Designer to calculate and predict the surface movement and deformation during subsequent filling mining. Industrial validation confirmed the feasibility of employing an artificial neural network to predict surface movement and deformation. The predicted results indicated maximum settlement values of -38.37 and − 39.73 mm in the strike and inclined main sections, respectively, with a prediction accuracy of 90.56%. The predicted settlement values were generally lower than the measured values. Therefore, correction coefficients of 1.26 and 1.40 are recommended for the strike and inclined main sections to enhance prediction accuracy.

Keywords

• Subsequent filling mining
• Surface movement and deformation
• Probability integral method
• MATLAB
• PCA-GA-BP neural network