Abstract: To accurately identify critical sections of the airport ground taxiing system and improve the congestion risk control capability，a traffic network model in the active area of the airport is built in this paper to mine critical nodes from four perspectives：node proximity，global path，feature vector and network location.Thus，an improved betweenness centrality algorithm is proposed to obtain a sample set of node importance ranking.The network performance indexes such as robustness，performance degradation rate and performance loss per unit time are integrated，and the performance change curve is analyzed by destructive experiments to determine the node importance ranking results of the complex network under the comprehensive destructive index CDI measure and achieve critical path identification.Finally，an airport in North China is taken as an example to conduct an empirical study.Our results show the traffic network in the active area of the airport has a small clustering coefficient and ismore resistant to destruction under random attacks，and the node sequence based on degree centrality is the most satisfying critical path in actual operation among all samples，providing references for controllers to implement traffic deployment on key sections and suppress conflicts.