Operational Data-Based Assessment of Air Traffic Controller Competence

A systematic evaluation of air traffic controller competence is crucial for enhancing airspace operational efficiency. This study develops an objective evaluation index system based on operational data, integrating radar track, controller, and meteorological information. Key indicators relevant to controller performance are identified using Pearson correlation and the random forest algorithm. Fuzzy C-means clustering is then applied to categorize controller ability levels, followed by sample balancing to improve data representativeness. A support vector machine model, optimized through grid search, is constructed to assess controller competence. The model’s reliability is validated using the F1 score and cross-validation, demonstrating effective and accurate identification of controller performance levels. This approach provides a quantitative and data-driven foundation for talent selection and airspace management improvement.