Adaptive Landing Sequencing using Bipartite Graphs and Perfect Matching

An optimal aircraft landing sequence is a sequence that results in the lowest runway engagement time, defined as the time the runway is in operation during the landing phase of flight. The approach presented in this paper uses a generic arrival fleet analyser, a cluster finding algorithm to isolate independent groups and a permutation enumeration algorithm to determine valid landing sequences. The optimal sequence of each cluster is determined and this result is used to develop a global optimal sequence for the arriving aircraft. The model was validated by simulating 24-hour arrival periods with varying arrival frequency distributions. The results indicate that, by exploiting the clustering effect, computational time can be conserved, and runway engagement time can be reduced by several minutes over a 24- hour period. This indicates a dynamic and adaptive sequencer is possible and through further development the model's accuracy and viability in a real environment can be validated.