There is a need for innovative technology to identify the filler material in recovered unexploded ordnance (UXO). The PELAN (Pulsed ELemental Analysis with Neutrons) system is being developed and tested, with support from the Environmental Security Technology Certification Program (ESTCP) and Naval Explosive Ordnance Disposal Technology Division (NAVEODTECHDIV), for non-intrusively and quickly identifying the filler of UXO in situ. PELAN interrogates a test object with neutrons and measures the resulting gamma ray spectrum to estimate the elemental composition of the sample material. The objective of this project was to investigate, test, and demonstrate advanced data analysis and decision making algorithms for the PELAN system.
In PELAN, the critical elements (carbon, nitrogen, oxygen, and hydrogen) are estimated using a linear signal model and a measure of the signal and background spectrum. A least squares fit is utilized to obtain estimates of the count rates of the various elements. This project investigated alternative signal models to more accurately reflect the underlying physics associated with the sensing modality, alternative spectral estimation procedures, and statistical algorithms to more effectively process the count data.
The researchers investigated several advanced data analysis algorithms and techniques to apply to PELAN data. Both the Least Squares/Generalized Likelihood Ratio Test (LS/GLRT) and the Least Squares/Principal Component Analysis (LS/PCA) combinations showed improvement over the LS/decision tree approach. As a result, the LS/GLRT method was implemented into the portable PELAN unit. The researchers continue investigation into the PCA spectral analysis method, which shows more improvement over the LS/GLRT approach. The PCA algorithm was shown to be effective at using the entire spectrum to extract characteristics of the target for improved identification. The advanced algorithms and processing techniques developed in this project and SERDP Project MR-1384 were considered in the conceptual design for ESTCP Project MR-200503.
The advanced data analysis and decision making algorithms for the PELAN system, developed under this project, will increase the filler detection efficiency and accuracy, reduce false alarm rates, and improve the system's ability to learn the signatures of new targets. These improvements will lead to cost savings and safer environmental remediation of UXO.