Optimum airspace design with air traffic controller workload-based partitioning.

摘要

This dissertation proposes an Air Traffic Controller (ATC) workload based methodology for optimum airspace partitioning. Initially, we define a set of airspace metrics for analytical modeling of the ATC cognitive workload and airspace complexity. We use a large-scale, fast-time simulation to model the current sectors in five Air Route Traffic Control Centers (ARTCCs) and compute the airspace metrics for each sector. These metrics are then used to calculate ATC workload and traffic complexity during various time intervals. Sectors are then ranked based on their traffic complexity and we show that the defined metrics are able to identify the complex sectors. Having a reasonable ATC workload modeling technique, we decompose the U.S. national airspace into three layers using altitude ranges based on operational levels of low, high, and ultra-high airspace. Each layer is further tiled into 2,566 hexagonal cells (hex-cells) with 24 nautical mile sides. These hex-cells are assumed to be finite elements of airspace and ATC workload is modeled for each hex-cell using various airspace metrics. We apply visualization techniques to analyze the spatial and temporal distribution of the controller workload and to identify congested periods of the U.S. National Airspace System (NAS).; Having the workload values for each hex-cell during the congested periods, we develop clustering algorithms using optimization theory to cluster hex-cells and partition the airspace to ARTCCs and sectors. We first partition the airspace to ARTCCs and define the optimum boundaries for different number of ARTCCs. Then the partitioning is continued within each ARTCC to construct optimum sector boundaries. This dissertation concentrates on simulation as a means to evaluate cognitive workload for the elements of airspace regardless of current sector and ARTCC boundaries. The only apriori inputs are the location of current ARTCC facilities and airports, the demand profiles for each city pair, and the filed routes. The proposed grid-based optimization methodology enables the inclusion of a wide range of objective functions and constraints.; This research should be of interest to both airspace design engineers and air transportation policy makers.