Abstract: To meet the air demand of working faces in underground mines, a method for adaptive regulation of branch airflow was proposed. A mathematical model of the underground ventilation network was established based on the 3 fundamental laws of mine ventilation. A nonlinear programming model was formulated with the minimum fan power as the objective function, and constraints were handled using the penalty function. The optimal set of adjustable branches and the airflow regulation range were determined based on ventilation network sensitivity analysis. The moss growth optimization algorithm was introduced to optimize the airflow optimization process, and its convergence performance for ventilation network solving was validated through performance comparison analysis. Based on the intelligent experimental platform of mine airflow regulation, the airflow regulation scheme for sensitive branches was experimentally verified. The overall system structure was designed; relevant functional modules were developed, and the IMGO algorithm was invoked to optimize the target airflow, outputting the optimal airflow regulation scheme and specific regulation parameters. The feasibility and effectiveness of the airflow regulation scheme based on sensitive branches were verified. The results show that the optimal fan power obtained using the IMGO algorithm is 240.56 kW; the average fan pressure is 2 588.53 Pa; the average number of iterations is 270, and the average iteration time is 20.5 s. All these data are significantly better than those of the other two optimization algorithms. Therefore, the improved moss growth optimization algorithm shows considerable improvement in both optimization performance and stability, enhancing the efficiency of solving underground ventilation networks. The research findings provide a reference for the optimization of mine ventilation networks.