NASA Logo, National Aeronautics and Space Administration

+NASA Home

+Ames Home

ATCorp's "Novel, Multidisciplinary Global Optimization under Uncertainty" Phase 2 Proposal With Co-I Nikunj Oza Awarded Funding
Intelligent Systems Division Banner

ATCorp's "Novel, Multidisciplinary Global Optimization under Uncertainty" Phase 2 Proposal With Co-I Nikunj Oza Awarded Funding

ATCorp won a one-year phase 2 award from the NASA Aeronautics Research Mission Directorate (ARMD) Leading Edge Aeronautics Research for NASA (LEARN) program for a project entitled, "Novel, Multidisciplinary Global Optimization under Uncertainty.” The goal of the project is to learn the best departure release times for commercial aircraft by utilizing a combination of simulations to represent the current scenario (traffic, weather, and other aspects), probabilistic networks to represent the distribution of possible future scenarios (and generate sample scenarios), and genetic algorithms to identify the best future scenarios (using criteria that can be specified). The tools are used to determine the flight pushback time that is most likely to yield the best future scenario.

As part of this project, Nikunj Oza will apply machine-learning methods for surrogate learning, which aims to learn from the inputs and outputs of simulations in a judicious way by running relatively few simulations. The hope is to get a machine-learning model that can be used in the future to get approximate simulation results much faster, and avoid running the more costly simulation where possible. Nikunj will also provide some general data mining expertise to the project.

BACKGROUND: In Phase I of this project, ATCorp’s team integrated two different technologies — Bayesian Networks (BNs) and Genetic Algorithms (GAs) — to develop a methodology called Probabilistic Robust Optimization of Complex Aeronautics Systems Technology (PROCAST) for global optimization under uncertainty. PROCAST is generally applicable to a wide range of complex problems displaying certain characteristics. One such problem is Integrated Arrival-Departure-Surface (IADS) air traffic management in busy, metropolitan areas.

In Phase I, the team produced an IADS traffic management tool for efficiently managing traffic on an airport surface and in the terminal airspace while maintaining robustness to unforeseen disturbances. They conducted proof-of-concept simulation experiments that showed that PROCAST, if applied at the John F. Kennedy (JFK) International Airport in New York City, would provide significant benefits over the current system. Encouraged by promising Phase I results, the Phase II project will work to expand and generalize the approach to work on problems of greater scope, complexity, and uncertainty, such as expanding to multi-airport scope and multi-objective cost functions that will include safety and other metrics.

PROGRAM FUNDING: Leading Edge Aeronautics Research for NASA (LEARN) program, NASA Aeronautics Research Institute (NARI), Aeronautics Research Mission Directorate (ARMD)

POC: Nikunj C. Oza,

First Gov logo
NASA Logo -