Real-Time Sub-Time Early Warning of Airport Scheduled Flight Delay Based on Immune Algorithm

摘要

The paper combines the characteristics of airport flight operation and statistical data of flight delays and forms early warning grading standards of flight delays. It applies the negative selection algorithm to diagnose and forecast the state of the airport scheduled flight delay, uses the probability of flight delay in a certain period as divisiory standards of mature detector and memory detector, presents the concept of the relative age in eliminating anile detectors in the process of training detectors. It uses corresponding flight data according to different time section of different date, increasing the week of flights weight coefficient on the basis of basic training flights data and improving the detecting and forecasting accuracy. The simulation experiment indicates, this method can veraciously forecast accumulative delay scheduled flight quantity in some time section, and carries on accurately early warning.