关键词: 航迹预测, 状态估计, 飞机意图, 性能模型, 环境模型, 机器学习

Abstract:

Aircraft trajectory prediction is the basis of air traffic management technologies such as flow management, conflict detection and resolution, aircraft approach sequencing, and abnormal behavior monitoring. There are many classic methods and application fields in the research of aircraft trajectory prediction. The background and significance of the research problem are summarized, and the basic knowledge of trajectory prediction is introduced from the database, basic process and key prediction technology. The databases include aircraft performance database, aircraft monitoring database and meteorological database. The basic processes include four modules of preparation, prediction, update, and output. The key prediction technologies summary and list typical models of three methods：state estimation model, dynamic model and machine learning model. When analyzing the models in detail, the main research results of the three methods are further listed and the characteristics of each method are summarized. The specific application of trajectory prediction in air traffic management is analyzed, including conflict detection, arrival management and flow management, etc. The challenges faced by the current trajectory prediction problem and the future development direction are summarized and pointed out.

Key words: trajectory prediction, state estimation, aircraft intent, performance model, environment model, machine learning