Through allocated methods of variable priority, a new arrival and departure scheduling optimization model based on the matching of the traffic flow with the capacity is established with an objective of minimum total delay time subject to constraints, i.e., the number of flights waiting in the airport, the airspace capacity, the surface capacity and the departure demand of the airport. A new selection method based on the cooperative co-evolutionary genetic algorithm is established using a pair of represents to form a cooperative group, which can effectively solve the low population diversity and premature of traditional genetic algorithm. The simulation results indicate that the model can effectively reduce the total delay time of the flights and meet the capacity limit requirement of the airport.