Aiming at mine ventilation network optimization problems, a hierarchy cultural algorithm is proposed by adapting
dual structure of cultural algorithm. In the population space, the hierarchy genetic algorithm is adapted. Hierarchy code method is introduced to describe the branch structure of ventilation network. The repair operator for infeasible branches and the regulation operator for feasible branches’ resistance are given. In the belief space, common dominating branches are extracted as knowledge by statistical learning method and used to guide the selection of adjustive branches in the evolution. Simulation results show that the optimal scheme obtained by the proposed algorithm can meet the needs of air control and has the minimum total resistance. That is, it has the less cost for air control than others.