Abstract: To address the issues of decision-making delay and insufficient parameter accuracy in traditional anti-wind control during mine fire accidents, a method for dynamic simulation of anti-wind effect in coal mine ventilation systems and construction of an airflow regulation parameter database was proposed. Taking the ventilation system of Sandaogou Coal Mine in Yulin City, Shaanxi Province as an example, a roadway topology database was established through 3D modeling technology. An improved genetic-ant colony hybrid algorithm was used to solve the ventilation network, achieving high-precision dynamic simulation of underground airflow distribution during fire periods. Through multi-scenario fire simulations, an anti-wind feasibility evaluation system was established, focusing on parameters such as smoke diffusion paths, variation coefficients of wind speed at key nodes, and anti-wind compliance time. Based on simulation data, a hierarchical airflow regulation parameter database was constructed, and an intelligent mapping of roadway numbering, fire coordinates, and regulation parameters was realized using associative coding technology. Practical application showed that the parameter database reduced anti-wind preparation time by 68% and decreased airflow stability standard deviation from ±15.3% to ±5.7%, enhancing the mine's emergency response efficiency to sudden incidents and providing a new paradigm for the application of intelligent ventilation systems in mine disaster emergencies.