A Data-Driven Framework for Operational Analysis and Traffic Pattern Identification in Multi-Airport Terminals

Abstract

The terminal airspace is considered the most complex area within the air traffic system, as it encompasses multiple nearby airports whose operations are interdependent, thereby increasing the complexity of management. A thorough understanding of airspace traffic patterns and operational characteristics is crucial for ensuring the safety and stability of air traffic. The paper proposes a data-driven analysis framework for traffic patterns in multi-airport terminal airspace. This framework utilizes machine learning methods applied to airport arrival and departure trajectory data to explore the operational characteristics within the terminal airspace. The framework comprises (i) a trajectory pattern identification module, which identifies trajectory patterns from a large volume of trajectories and analyzes the characteristics of these patterns, and (ii) a traffic flow pattern identification module, which utilizes the trajectory pattern to identify traffic flow patterns within the terminal airspace, thereby characterizing the operational structure of airspace traffic flows and the spatiotemporal dependency between trajectory patterns. This framework can analyze the operational characteristics of trajectory patterns, route intersections, and traffic flow patterns within the airport terminal area. It helps managers better understand aircraft behavior in the terminal area, identify risk locations at intersections in the airspace, and reveal typical traffic flow structures. This supports the optimization of terminal area airspace structure and provides decision-making tools. By analyzing the terminal airspace operations of two airports in Shanghai (ZSPD and ZSSS), the framework’s outcomes and capabilities are demonstrated. The study found that the airspace design of ZSPD is relatively complex, identifying multiple prevalent trajectory patterns. It also revealed that the airspace traffic flow structure exhibits certain temporal regularities, which aids in predicting the airspace’s operational structure and capacity, thereby facilitating informed decision-making.