Аннотация
Fall of Ground (FoG) incidents remain among the most critical safety and operational hazards in underground mining. These events, including roof falls, rib spalling, and localized collapses—typically result from complex interactions between stress redistribution and structural weaknesses in the rock mass. Every year, mines worldwide continue to report FoG-related casualties despite advancements in ground support technologies. A quantitative assessment of FoG hazards is therefore essential to support the implementation of effective and reliable ground control strategies. This study introduces a causality-based Rock Engineering System (RES) for the risk assessment and prediction of FoG hazards in underground excavations. A FoG Risk Index is proposed to quantify the hazard level systematically and evaluate the influence of contributing factors. Traditional FoG assessment methods often rely heavily on qualitative judgment or are limited in their ability to handle complex interactions among geological, structural, and stress-related factors. Moreover, they frequently lack mechanisms to incorporate uncertainty and the reliability of expert input. To address these limitations, an integrated and data-driven approach that captures both the causal relationships among factors and the inherent uncertainty in expert knowledge is needed. The RES is weighted using a fuzzy cognitive map (FCM), with uncertainty and expert judgment reliability incorporated through Z-number theory—resulting in a robust framework for FoG hazard prediction. A database of 150 FoG cases from a local underground mine was compiled to develop and validate the model. The results indicate that the predicted FoG Risk Index aligns well with observed FoG events, indicating the model’s accuracy and practical applicability. It is concluded that the proposed index can be effectively used to assess FoG risk in underground excavations.