Integrating Rule-Based and Neural-Net Techniques for Spectral Analysis
Abstract
Spectral analysis involving the determination of atomic and molecular species present in multi-spectral data is a very time consuming task, especially considering the fact that there are typically thousands of spectra collected during each experiment. Due to the overwhelming amount of available spectral data and the time required to analyze these data, a robust automatic method for performing preliminary spectral analysis is needed. This research focused on the development of a rule-based expert system and a supervised artificial neural network with error correction learning, specifically a three-layer, feed- forward, back-propagation perceptron. The objective was to develop an integrated spectral analysis system which would perform preliminary spectral analysis and save the analysts from the task of reviewing thousands of spectral frames. The input to the neural network, which is screened by the rule-base, is raw spectral data, with the output consisting of the classification of both atomic and molecular species in the source.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 1990
- Accession Number
- ADA223517
Entities
People
- Arthur L. Sumner
Organizations
- Air Force Institute of Technology