LAUSR

To understand the mechanism of delay propagation from the perspective of multiple airports, constructing delay propagation relation (DPR) networks among airports is a novel analysis method. The latest method is to use transfer entropy to mine the delay propagation relation among airports. However, the transfer entropy will produce bias due to estimating high-dimensional conditional mutual information (CMI) in the delay propagation scenario. In this paper, we propose the low-dimensional approximation of CMI for transfer entropy (LTE) to address the above issue. By applying this improved algorithm, the delay propagation relation among airports can be more accurately explored and a more accurate DPR network can be obtained. For this network, we apply the complex network theory and its related indicators to provide systematic analysis about delay propagation. The results of case analysis show that in the delay propagation interactions among airports, large airports always receive delays and small airports propagate delays outward. Meanwhile, delays propagate more efficiently in the aviation system of smaller airlines. These results can provide some theoretical supports for making measures to reduce delay propagation.