Airport Cascading Failure Modeling with Improved Coupled Map Lattice and Its Application to Robustness Analysis of Airport Networks

Abstract

“Cascading failure” is a critical concept for characterizing the structures and dynamics of complex networks, which describes the chain reaction triggered by a single failure or error in a system or process. In the field of airport networks, cascading failure refers to the airport cascading failure (ACF). In recent years, most studies on ACF only focused on the state changes of airport nodes with the cascading failure models and neglected the impacts of propagation of flight delay (PFD). To analyze the robustness of airports in the sense of ACF, the improved coupled map lattice (I-CML) model is proposed by combining the impacts of PFD and the throughput of the airport. As an illustration, the Chinese airport network (CAN) is classified into four categories with the K-means clustering algorithm and the airport classification indicator. Our exploration discovers two interesting results: firstly, the I-CML model is more robust than the original CML model when identical attacks exist; secondly, the critical nodes of CAN are more likely to cause cascading failures and the important node attack can be modeled effectively by the random attack. In the perspective of airport network operations, it is necessary to protect the critical airport nodes to ensure the stability of the airport network.