Abstract: Through a comprehensive review and analysis of research on the migration law of airflow catastrophe in mine roadways, optimal sensor placement, ventilation system fault diagnosis, and ventilation control strategies for ventilation networks, this study proposes a source-term non-stationary ventilation network calculation model for different stages of gas emission effect and gas pressure effect. The model provides data support for the catastrophe state identification of mine roadway airflow and the decision-making of ventilation control strategies. By combining meta-heuristic optimization algorithms with machine learning, an optimal sensor placement scheme suitable for catastrophic airflow monitoring is generated, and a dynamic identification model for airflow catastrophe state is constructed. This model realizes the combined identification of air volume and gas concentration as well as multi-point resistance change diagnosis and identification. Through the dynamic integration of the non-stationary ventilation network calculation model and ventilation network adjustment methods, a dynamic regulation model for non-stationary ventilation networks is established, including ventilator regulation, branch air resistance regulation, and combined regulation of ventilator and air resistance. This model enables simulation analysis of collaborative ventilation control for ventilation networks during disasters. Taking reducing the disaster-affected area, preventing branch airflow reversal, and rapidly exhausting high-concentration gas as the ventilation control criteria, a segmented hybrid multi-objective optimization method based on conditional distribution sampling is proposed to solve the dynamic regulation model, providing means and basis for the decision-making of the optimal ventilation control scheme during disasters. Finally, an experimental system for monitoring, identification and collaborative ventilation control of airflow catastrophe state in mine roadways is built for verification, which provides theoretical guidance for solving the scientific challenge of identifying and controlling airflow catastrophe states during disasters.