Abstract: Aiming at the problem of large-scale flight delays caused by insufficient airspace resources under severe weather conditions, this paper proposes a repair optimization strategy for air route networks based on the complex network repair theory and traffic flow distribution theory and drawing on the traffic network design theory. The repair scenario of the air route network is first established, and the flight constrained area in severe weather generated based on weather information. A bi-level programming repair model is then constructed, in which the upper model takes the lowest repair cost as the objective function, and the lower model is the multi-constrained traffic flow distribution model. The improved particle swarm optimization algorithm and the K shortest path algorithm are applied to solving the whole model and the lower model, respectively. Finally, two types of indicators are proposed to evaluate the repair effect of the air route network from both the local and global perspectives. Based on the typical route network, two basic repair strategies are compared, and the actual operation results compared to study the repair effect and applicability of different repair strategies. The simulation results show that the air route network repair optimization strategy can not only make up for the structural limitation of the repair strategy based on original topology, but also solve the problem of the additional coordination cost brought by the repair strategy based on adjustment topology. The proposed repair optimization strategy minimizes the disruption to normal flight operations and restores all affected flights to normal operation with minimum repair costs, exhibiting considerable significance in mitigating route congestion and flight delays.

Key words: air route network repair, complex networks, severe weather, bi-level programming models, particle swarm optimization, K-shortest path algorithm