NICOP - Analytical and Numerical Methods for Identifying Principal Components of Heterogeneous Physical Data

Abstract

Analytical and Numerical Methods for Identifying PrincipalComponents of Heterogeneous Physical DataThis proposal is submitted in response to ONR Broad Agency Announcement (BAA) #N00014-17-S-BA001 ~Long Range Broad Agency Announcement (BAA) for Navy and Marine CorpsScience and Technology~ published on 1 October 2016.Project Summary:The objective of the proposed NICOP project is the develo"pment and assessment of analyticaland numerical methods for identifying principal components from heterogeneous physical data,for"" application in ship hydrodynamics, hydroelasticity, and naval systems. The long-termobjective is the establishment of rigorous met""hods for a better assessment, understanding,representation, dimensionality reduction, and prediction of the physics of complex phen""omena,such as those involved in the operation of military ships, subject to a variety of environmentaland operating conditions. Th""e approach will include linear and nonlinear methods for distributedand lumped physical data assessment, variability breakdown, and"" dimensionality reduction.Linear methods include principal component analysis techniques, whereas nonlinear methodsinclude kernel"" and local principal components analysis combined with clustering techniques, aswell as deep autoencoders. Potential problems of in""terest include physical space assessment forcomplex hydrodynamics and hydroelasticity, dimensionality reduction of simulation ande""xperimental data, extended validation of simulations versus experiments by principalcomponents. The peculiarities of the methods pr""oposed (aimed at an augmented physical spaceassessment) constitutes an original contribution to the ship hydrodynamics field, meeti"ng thegoals of the ~Platform Design and Survivability~ US Naval S&T Focus Area.Grantee:Principle Investigator:PI NAME: Dr. Matteo DiezDept. of University/Organization: CNR-INSEAN (National Research Council-MarineTechnology Research Institute)Address: Via di" Vallerano 139, Rome, ItalyEmail: matteo.diez@cnr.it Phone: +39 06 50299 314 Fax: +39 06 5070 619Administrative POC:Name: Serenel"la VitaliTitle: Administrative secretaryDept. of University/Organization: CNR- INSEAN (National Research Council-MarineTechnology" Research Institute)Address: Via di Vallerano 139, Rome, ItalyEmail: serenella.vitali@cnr.it Phone: +39 06 50299 291 Fax: +39 06 5"070 619

Document Details

Document Type
DoD Grant Award
Publication Date
Dec 20, 2017
Source ID
N629091812033

Entities

People

  • Matteo Diez

Organizations

  • Office of Naval Research
  • United States Navy

Tags

Readers

  • Agent-Based Social Robotics and Mobile-Assisted Learning in Virtual Environments.
  • Distributed Systems and Data Platform Development
  • Technical Research and Report Writing.

Technology Areas

  • Space