Dynamic probabilistic risk assessment of emergency response for intelligent coal mining face system, case study: Gas overrun scenario

Intelligent coal mining has become an inevitable development trend of the coal industry in the era of artificial intelligence. Emergency response level is particularly important for preventing accidents and reducing accident losses in intelligent coal mine (ICM) operations, which urgently needs an accurate risk assessment method to assess the risk of ICM emergency response. Combining the ICM system structure and emergency response process (ERP) of human-system interaction based on the information, decision, and action in crew context (IDAC) model, this study proposes a dynamic probabilistic risk assessment (PRA) method based on dynamic Bayesian network (DBN), mainly including the event sequence diagram (ESD) from the cognitive perspective, human-system concurrent task analysis (CoTA) from success perspective and fault tree analysis (FTA) model from failure perspective, dependency model of performance influencing factor (PIF) from organizational impact perspective, and the “ESD-FTA-PIFs” coupled BN model. To address the uncertainty of risk factors (e.g., human factors and system equipment) and achieve quantitative analysis, the coupled BN model is mapped to the DBN model to perform dynamic PRA combining with existing databases, fuzzy set theory, Dempster-Shafer (D-S) evidence theory, and time-variant model. Finally, the practicality and advantages of the proposed methodology are demonstrated by a real case study in the gas overrun scenario. The results show that the DBN model can perform dynamic inference analysis to achieve the risk prediction of event and system state compared with the static BN model. The comprehensive integrated DBN model of “ESD-FTA-PIFs” not only enables bidirectional risk analysis of emergency response event, but also enables the systematic traceability of risk factors. The proposed methodology can be a useful tool to assess the dynamic risk of ICM emergency response and decision-making, and it can provide scientific guidance for risk assessment during emergency response of other safety-critical industries.