关键词: 交通流分布, 终端区空域, 多特征融合, 平均交通流

Abstract: Aiming at the problems that the airspace operation environment in the terminal area is complex and standard routes are difficult to dynamically reflect the actual air traffic control (ATC) command scenarios, this paper pro鄄 poses a terminal area traffic flow distribution identification method based on multi-feature fusion to accurately describe the real traffic situation and effectively evaluate the complexity of ATC. This method aims to solve the problems of uneven trajectory data length, single feature dimension and insufficient accuracy of traditional clus鄄 tering algorithms. Through resampling, the original trajectories are standardized into equal-length sequences; by fusing position, heading and velocity features, a similarity model with multi-feature fusion is constructed. Com鄄 bined with the improved hierarchical clustering algorithm, the average traffic flow is extracted to construct the terminal area route topology map under actual operating conditions. Verified by the arrival and departure data from Kunming Changshui International Airport, the results show that this method can effectively identify the traf鄄 fic flow distribution at each corridor entrance. The results obtained can better reflect the characteristics of dy鄄 namic ATC guidance than the standard aeronautical chart, which can provide a basis for the refined management of the terminal area and the analysis of ATC behaviors.

Key words:  , traffic flow distribution, terminal area airspace, multi-feature fusion, average traffic flow