An intelligent water source discrimination method for water inrushes from coal seam roofs in the Inner Mongolia-Shaanxi border region

Water hazard on the coal seam proof induced by high-intensity coal mining are increasingly prominent in the Inner Mongolia-Shaanxi border region. The effective, accurate water-source discrimination of the water inrushes is the key to water hazard prevention. This study investigated three typical mines in the Inner Mongolia-Shaanxi border region. To this end, principal component analysis (PCA) was employed to extract principal components from 80 groups of groundwater samples. Then, with inorganic indicators K⁺+Na⁺, Ca²⁺, Mg²⁺, Cl⁻, SO₄ ²⁻, HCO₃ ⁻ and TDS and organic indicators UV₂₅₄, TOC, and dissolved organic matter (DOM)’s fluorescence spectra as discriminant indicators, this study proposed a intelligent identificaton method of PCA-AFSA-RF roof water inrush source by using artificial fish swarm algorithm (AFSA) to improve random forest (RF). First, a PCA-RF discriminant model was established, with accuracy (A_c), precision (P_r), recall (R_c), and F-measure (f₁) of 83.00%, 83.17%, 80.42%, and 79.57%, respectively. Then, in the PCA-RF discriminant model, AFSA was employed to optimize the number of decision trees, the depth of trees, and the minimum sample number needed for internal node splitting. Furthermore, a genetic mechanism was introduced into AFSA to avoid local optimization. In this way, a PCA-AFSA-RF water-source discriminant model for water inrushes on coal seam roofs was established, with A_c, P_r, R_c, and f₁ of up to 92.18%, 91.11%, 87.58%, and 88.82%, respectively, increasing by 9.18%, 7.94%, 7.16%, and 9.25% compared to the PCA-RF model. Furthermore, the PCA-AFSA-RF exhibited a back substitution accuracy reaching 97.50%. Finally, this model was used for the water-source discrimination of 12 water samples from the mines, yielding results consistent with the actual results in the field. This indicates that the PCA-RF model with improved AFSA enjoys better accuracy and generalization ability. The research results of this study can provide a new method for the accurate water-source identification of water inrushes from coal seam roofs.