When and how to partition airspace: a data-driven control framework for dynamic airspace sectorization based on...

Area of Science:

• Air Traffic Management
• Operations Research
• Computer Science

Background:

• Dynamic airspace sectorization (DAS) is crucial for adapting to evolving air traffic and weather conditions.
• Existing methods face challenges in determining optimal sectorization timing and configuration.
• Current airspace management requires enhanced flexibility and efficiency.

Purpose of the Study:

• To develop a data-driven control framework for dynamic airspace sectorization.
• To address the 'when-to-do' and 'how-to-do' aspects of airspace re-sectorization.
• To improve the adaptability and efficiency of air traffic management systems.

Main Methods:

• Developed a knowledge construction module with sector generation, controller workload, and sector similarity models.
• Constructed a multi-objective receding horizon optimization (RHO) model for sectorization scheme generation.
• Implemented a decision execution module using multi-criteria decision analysis for re-sectorization timing and effectiveness analysis.
• Utilized Singapore Flight Information Regions (FIRs) trajectory data for empirical assessment.

Main Results:

• The proposed framework generates serialized optimal airspace sectorization schemes with high similarity.
• Compared to single interval optimization (SIO), the RHO model offers improved sector shape adaptation.
• The decision execution module effectively determines optimal re-sectorization timing based on future traffic.
• Empirical analysis validated the framework's effectiveness using real-world flight data.

Conclusions:

• The data-driven framework enhances dynamic airspace sectorization by optimizing both scheme generation and timing.
• The approach provides a more flexible and efficient solution for modern air traffic management.
• This methodology offers a significant improvement over traditional single interval optimization methods.