Flight alternate optimization scheme in dangerous weather based on multiexpectation

Abstract

Aircraft landing with low fuel is usually caused by diversion decision changes in the air. It could lead to an unsafe event. To solve the problem of collective alternate landings of multiple flights in the area, the most complex situation is selected, that is, when the weather in the route or terminal area is dangerous. However, in this situation, when the pilot announces diversion, many are unacceptable by the airport due to limited aircraft stand. Therefore, this study establishes a decision-making approach to help air traffic controllers and airlines choose suitable alternate airports since accurate fly limit zone conditions can keep the aircraft out of dangerous weather and a short diversion route can avoid the low fuel situation, both of which can increase the safety of the diversion process. The basic conditions of the fly limit zone under dangerous weather are obtained by combining meteorological data and historic tracks. Here, A* and the improved gray wolf Dijkstra algorithm are used to plan the diversion path for two different routes to the alternate aerodrome: maneuvering flight and flying along the route. First, considering the shortest total flight time of alternate flight as the single objective and subsequently integrating the expectations of flight, control, airport, and airline, a multi-objective function is constructed, the dynamic decision-making time interval is defined, and a dynamic optimization scheme of multi-flight alternate in the region based on a single objective and multi-objective is proposed. The single-objective scheme focuses on the safe aspect of diversion, while the multi-objective scheme focuses on preventing airport rejection. The multi-objective scheme can enable the flight to land quickly, ensure safety, and consider the expectations of multiple parties to avoid diverting simultaneously. Using the “8.12” North China large area alternate landing data, the total flight time obtained by selecting the direct flight A* algorithm is reduced by 100 and 62 min, respectively, and the total flight time obtained by the improved gray wolf Dijkstra algorithm based on the route is reduced by 73 and 14 min. Moreover, in the multi-objective scheme, the overall time of flight resumption to the original destination is 63 min earlier, and the total cost is reduced by 62900 yuan. Although multiple diversion still occurs on CA991, the process is within the safety range. Therefore, results indicate that the scheme considers the needs of multiple parties and improves the economy to ensure the safety of flight alternate landing.

Keywords

• air transportation
• dynamic diversion optimization
• improved gray wolf algorithm
• multi-flight diversion
• flight restriction area