Abstract: Aiming at the existing defects of fixed parameters, unstable convergence and others in the traditional intelligent algorithm in mine ventilation network calculation, the research integrates parameter self-adaptation mechanism and multi-strategy collaborative optimization theory, and proposes an improved intelligent iterative algorithm system; A dynamic mapping relationship between algorithm parameters and network topological features is established, and an intelligent balance between global exploration and local development is achieved through quantitative evaluation of population diversity and real-time diagnosis of convergence status; A population initialization method based on domain knowledge is designed, non-uniform arithmetic crossover and multi-mode mutation operators are adopted, and dynamic topological structures and domain search strategies are developed. In the verification of the complex ventilation network in Yushupo Coal Mine, Ningwu, Shanxi Province, the average number of iterations of the improved algorithm is reduced to 180 times, the total unbalanced flow rate of nodes is controlled at 0.048 m³/s, and the success rate of 30 times of independent operations is 100%. The algorithm parameter self-adaptation adjustment mechanism effectively solves the problem of fixed parameter settings being difficult to adapt to the dynamic changes of complex networks, and multi-strategy collaborative optimization significantly improves the efficiency and reliability of ventilation network calculation.