Abstract: Real-time acquisition of global airflow parameters is a crucial technology for the intelligent control of metal mine ventilation systems. This paper presents a method for sensing global airflow parameters in metal mines, aimed at providing the real-time airflow data necessary for issuing dynamic control instructions to intelligent control systems. Utilizing deep learning with multilayer perceptron algorithms, an AI model for mine ventilation networks was constructed. The method involves applying a three-dimensional ventilation simulation system to model airflow parameters under different fan operating conditions and natural wind pressure states, thereby establishing a training and testing dataset for the AI model. This approach forms a method for sensing mine airflow parameters that uses sensor monitoring data as input for the AI model and provides global airflow parameters as output, achieving real-time intelligent perception of global airflow parameters.Using a specific metal mine as a case study, the results indicate that the intelligent sensing model for mine airflow parameters demonstrates high goodness-of-fit and sensing accuracy, with a coefficient of determination (R2) of 0.998, a root mean square error (RMSE) of 0.2159, and a mean absolute error (MAE) of 0.085. This shows that by providing real-time monitoring data from a limited number of sensors to the AI model, reliable global airflow parameters can be obtained. This supports the formulation of intelligent control strategies for mine ventilation systems, effectively addressing the key technical challenges of real-time global airflow parameter acquisition and significantly contributing to the development of intelligent ventilation systems in metal mines.