A chance-constrained programming model for airport ground movement optimisation with taxi time uncertainties

Airport ground movement remains a major bottleneck for air traffic management. Existing approaches have developed several routing allocation methods to address this problem, in which the taxi time traversing each segment of the taxiways is fixed. However, taxi time is typically difficult to estimate in advance, since its uncertainties are inherent in the airport ground movement optimisation due to various unmodelled and unpredictable factors. To address the optimisation of taxi time under uncertainty, we introduce a chance-constrained programming model with sample approximation, in which a set of scenarios is generated in accordance with taxi time distributions. A modified sequential quickest path searching algorithm with local heuristic is then designed to minimise the entire taxi time. Working with real-world data at an international airport, we compare our proposed method with the state-of-the-art algorithms. Extensive simulations indicate that our proposed method efficiently allocates routes with smaller taxiing time, as well as fewer aircraft stops during the taxiing process