Study On Throughput Oriented En-route Sector Categorization

Accurate evaluation of sector capacity is the basis of efficient air traffic flow management. Recent research mainly concentrates on the evaluation of static sector capacity, while less research focuses on dynamic capacity. Most studies on influencing factors of sector dynamic capacity are primarily qualitative research. Therefore, quantitative research on influencing factors of sector dynamic capacity is of certain help of assessment of dynamic capacity.Through the analysis of definition of dynamic sector capacity, the influencing factors of sector dynamic capacity are identified by taking influencing factors of controller workload into consideration. Principal component analysis and PSO-Kmeans clustering are used to classify sectors. Describing the relationship of cluster and influencing factors lay the foundation of dynamic capacity evaluation.First, influencing factors of controller workload are analyzed and summarized from the sector traffic characteristics and controller performance. And weighted grey correlation degree analysis is conducted.Second, according to the definition of dynamic sector capacity, influencing factors of sector capacity are analyzed and summarized from the following aspect, available airspace resource, air traffic flow and control service. Influencing factors are classified into static factors and dynamic factors. Dynamic influencing factors are extracted to be as the sector classification indices. System of influencing factor index is established and analyzed using the principal component analysis method. By explaining the main decomposition of sample data, the significance of every index is identified.Finally, taking 10 sectors of Beijing area control center as an example, PSO-Kmeans clustering is performed based on principal component scores. The Silhouette Coefficient is calculated to identify the best K value of the cluster and to verify the validity of the cluster is verified. Moreover, clustering results are regarded as sector classification results which are related to controller workload.