关键词: 交通运输工程, 航空运输, 改进聚类算法, 禁忌粒子群算法, 飞行轨迹, 轨迹压缩, 模糊C-均值聚类

Abstract: In order to accurately grasp the flight mode of aircraft in terminal area and effectively evaluate and optimize flight procedures, firstly, according to the spatio-temporal characteristics of flight trajectory points, a top-down algorithm based on time ratio was proposed to compress the flight path. Secondly, combined with the velocity and heading characteristics of the trajectory points, the trajectory similarity model based on multi-dimensional features was established. Finally, the tabu search particle swarm optimization (TSPSO) algorithm was applied to improve and optimize the fuzzy C-means clustering (FCM) algorithm, and the improved clustering algorithm was verified by combining with the real flight trajectory data of the terminal area. The results show that the trajectory compression technology greatly reduces the computational cost. Compared with the traditional FCM algorithm, the improved clustering algorithm can obtain a better satisfactory solution and improve the flight trajectory clustering effect.

Key words: traffic and transportation engineering, air transportation, improved clustering algorithm, the tabu search particle swarm optimization (TSPSO) algorithm, flight trajectory, trajectory compression, the fuzzy C-means clustering