Modelling the delay propagation effects of multiple resource connections in an airline network using a Bayesian network model

Abstract

We developed an airline network delay propagation model using a Bayesian Network in a delay-tree framework, accounting for multiple connecting sources of an airline including aircraft, crew connections, and passenger connections. A case study was conducted to examine the heterogeneous delay propagation effects in a flight network. Results showed that the model was able to identify weak links in a flight network based on past operational data. Two new delay multiplier metrics were developed and test results showed that the Expected Delay Multiplier could better account for historical flight delay profiles by inferring posterior probability distributions based on new observations.