| In this work,based on the uncertainty theory,the flight normality is predicted which is a hot topic in current academic research.Planned delays refer to a situation in which delays are caused by insufficient flight schedules at various stages,such as flying,turnaround.First,this work proposed a distributed modeling method based on Bayesian theory,which lays a foundation for the modeling of each link of flight operation.In the process of modeling,cubic spline interpolation is also introduced to solve the problem of distribution multiplication.Then,the uncertain distribution model of each link is established according to dividing flight operation into four links: taxi-out,flying in the air,taxi-in and turnaround.Based on this,the uncertain distribution of the whole flight operation process is calculated by using the serial uncertain event superposition algorithm,which provides a basis for estimating the planned delay caused by inadequate stops.In addition,this work calculates the serial superposition effect for each flight chain to characterize the ripple effect of flight planned delays.Finally,by abstracting the flight into a process and the aircraft into a core,this paper describes the flight operation process in a parallel system way.The robustness of the flight schedule is also studied.Experiments show that the method proposed in this work can effectively predict planned delays from the perspective of single flight,flight chain and flight schedule.To that end,this paper can provide decision-making basis for airlines to make reasonable flight plans. |
