You are here

NonLinear Parallel OPtimization Tool

Award Information
Agency: National Aeronautics and Space Administration
Branch: N/A
Contract: NNX17CG08C
Agency Tracking Number: 155690
Amount: $749,968.00
Phase: Phase II
Program: SBIR
Solicitation Topic Code: H9.03
Solicitation Number: N/A
Timeline
Solicitation Year: 2016
Award Year: 2017
Award Start Date (Proposal Award Date): 2017-04-10
Award End Date (Contract End Date): 2019-04-09
Small Business Information
301 North Neil Street, Suite 502
Champaign, IL 61820-3169
United States
DUNS: N/A
HUBZone Owned: No
Woman Owned: No
Socially and Economically Disadvantaged: No
Principal Investigator
 Alexander Ghosh
 Principal Investigator
 (217) 721-2875
 ghosh@cuaerospace.com
Business Contact
 David Carroll
Title: Business Official
Phone: (217) 239-1703
Email: carroll@cuaerospace.com
Research Institution
N/A
Abstract

The technological advancement proposed is a novel large-scale Noninear Parallel OPtimization Tool (NLPAROPT). This software package will eliminate the computational bottleneck suffered by many standard NASA-utilized analysis tools such as GMAT, EMTG and NASTRAN. Currently these programs rely on serial nonlinear programming solvers such as the Sparse Nonlinear OPTimizer (SNOPT), despite the fact that their own codebases support full parallelization. The same is true for tools used in other industries for applications such as electric power grid optimization, nuclear reactor control and stock market analysis. The NLPAROPT algorithm can be quickly incorporated into these existing software solutions via a user-friendly interface and will offer an instant runtime reduction for very large-scale optimization problems. Irrespective of runtime gains, Phase I analysis has shown that the NLPAROPT algorithm is capable of outperforming industry standard serial solvers such as SNOPT for tested problems, including complex trajectory design problems. The Phase I effort has also identified several potential computational research avenues that, once completed in Phase II, will result in massive execution speed increases, further improving the attractiveness of this new parallel algorithm.

* Information listed above is at the time of submission. *

US Flag An Official Website of the United States Government