Advanced TFM Congestion Management Performance Analysis and Research Results

Abstract

This document describes the results of Fiscal Year 2009 (FY09) research on advanced congestion prediction and automated en route congestion resolution. An improved model for aggregate traffic demand prediction uncertainty was completed. Three different models for estimating the impact of weather on sector capacity were compared, and a hybrid solution is proposed. A new technique for developing rerouting options was developed for application to near-term, semi-automated congestion management tools. An existing sequential decision-tree approach for tactical, probabilistic congestion management was converted to a continual approach that can be realistically applied to real-time decision support. Finally, two methods for improving the performance of automation-developed congestion resolution maneuvers were studied: a partial-optimization approach that can consider multiple congestion resolution goals, and an approach for adapting to poor forecasts by explicitly planning deferred resolution maneuvers.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 01, 2009
Accession Number
AD1108352

Entities

People

  • Anthony J. Masalonis
  • Christine P. Taylor
  • Claude K. Jackson
  • Craig R. Wanke
  • Lixia Song
  • Norma J. Taber
  • Stephen M. Zobell

Organizations

  • MITRE Corporation

Tags

Communities of Interest

  • Human Systems
  • Space

DTIC Thesaurus Topics

  • Accuracy
  • Air Traffic
  • Air Traffic Control Systems
  • Algorithms
  • Computational Science
  • Control Systems
  • Correlation Analysis
  • Data Science
  • Databases
  • Genetic Algorithms
  • Heuristic Methods
  • High Altitude
  • Information Science
  • Knowledge Management
  • Mathematical Models
  • Monte Carlo Method
  • Navigation
  • Probability Distributions
  • Regression Analysis
  • Three Dimensional
  • Two Dimensional
  • Weather Forecasting

Readers

  • Computational Modeling and Simulation
  • Computer Networking
  • Computer Vision.