Novel Methods For Rapid, Reliable, And Rigorous Analysis Of Gamma-Ray Spectra Using Optimization And Nuclide Modeling
Abstract
This research will develop and implement a unique combination of state of the art mathematical methods for spectrum analysis using nuclide buildup and decay modeling, convex optimization, and Bayesian statistics. The analysis engine we develop will automatically identify gamma-emitting radionuclides within the sample, resolve interferences, and simultaneously calculate activities of entire decay chains in a single analytic step. This approach to spectral analysis will increase the quality of post-detonation response technology by minimizing variance among analysts and increase confidence in results, apply a rigorous statistical treatment of all uncertainties, and give a complete report of results including probability distributions for nuclide attribution and activity calculation, as well as covariances, mean values, and standard uncertainties for all results. Phase I will focus on the development of the algorithms and subroutines for the new approach to spectrum analysis as well as the design and collection of spectral data that will span a range of detector types and measurement parameters expected in post-detonation response scenarios and standard gamma-ray spectrum analysis capabilities. Phase II will focus on the analysis of the test datasets using the new methodology as well as traditionally employed techniques as a basis for comparison
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Oct 04, 2018
- Source ID
- HDTRA11710055
Entities
People
- Steven Biegalski
Organizations
- Defense Threat Reduction Agency
- Georgia Tech Research Corporation