A Robust Approach for Predicting Dynamic Density

This research proposes and analyzes an approach for predicting controller workload by predicting dynamic density. Most dynamic density formulations estimate workload with a linear combination of a set of dynamic density factors that describe the trac situation in a sector. The robust approach proposed here uses this linear structure and the available data to explicitly consider the relative levels of uncertainty in dynamic density factor predictions when predicting dynamic density. The benets of the robust approach are analyzed by using predicted and actual dynamic density factor data collected while playing back trac data in the Future ATM Concepts Evaluation Tool. Results indicate that the robust approach produces errors that are more than an order of magnitude smaller than those produced by a simple approach that ignores factor prediction uncertainties. However, other approaches achieve lower prediction errors than the proposed robust approach.