Dynamic airspace sectorization (DAS) is a core technology in airspace management, aimed at addressing continuously changing traffic patterns and unpredictable severe weather. Regarding the two fundamental issues of when-to-do and how-to-do a sectorization, a data-driven control framework is developed in this paper, which includes two key modules: the knowledge construction module and the decision execution module. The knowledge construction module mainly defines the knowledge related to airspace sectorization problem, including sector generation model, controller workload model, and sector similarity model. Based on these underlying components, a multi-objective receding horizon optimization (RHO) model is constructed to generate sectorization schemes with gradually changing shapes, solving the task of how-to-do a sectorization. The decision execution module introduces multi-criteria decision analysis and automated decision logic to solve the task of when-to-do a re-sectorization, including selecting an optimal sectorization scheme and analyzing its effectiveness for future traffic situations. Using trajectory data from the Singapore Flight Information Regions (FIRs), a comprehensive assessment and analysis of the two modules is conducted. Empirical findings indicate that, compared with the single interval optimization (SIO) approach, our framework can generate serialized optimal airspace sectorization schemes with high similarity, which will enhance the flexibility of future airspace management.
Du et al. (Wed,) studied this question.