Revista ELECTRO
Vol. 46 – Año 2024
Artículo
TÍTULO
Identificación de Radioisotopos en Tiempo Real Aplicando Mínimos Cuadrados Usando C++ para una Arquitectura Embebida
AUTORES
Pacheco-González, A.; Rivera, M.; Chacón-Blanco, R.; Robledo-Vega, Isidro
RESUMEN
Se presenta una implementación eficiente de un método de análisis espectral de rayos gamma de baja resolución aplicando mínimos cuadrados no-negativos para identificar radioisótopos en tiempo real en una plataforma embebida por medio de una librería de C++ altamente optimizada para procesadores ARM 64 utilizando el entorno de desarrollo Xcode 15 para iOS 17 mezclando códigos escritos en lenguaje C++ 17 dentro del lenguaje nativo Swift, explotando las capacidades propias de la arquitectura de compilación LLVM. Se alcanzaron tiempos de ejecución del algoritmo RIID inferiores a los reportados en la literatura. Se discuten posibles mejoras tanto del algoritmo como del conjunto de datos de prueba.
Palabras Clave: Identificación de radioisótopos, análisis espectral de rayos gamma, mínimos cuadrados, sistema embebido.
ABSTRACT
A very fast embedded software implementation of a low-resolution gamma ray spectral analysis method using non-negative least squares for real-time radioisotope identification is presented. This implementation was carried out using a highly optimized C++ library for ARM64 processors using the Xcode 15 development environment under iOS 17, mixing C++ 17 code with native Swift language, exploiting the capabilities offered by LLVM compilation architecture. Average execution times of the RIID algorithm were lower than those reported in the literature. Possible improvements to both the algorithm and the testing datasets are discussed.
Keywords: radioisotope identification (RIID), non-negative least squares, gamma ray spectral analysis, embedded system.
CONTENIDO
REFERENCIAS
[1] U. Yoshinori et al, Surveillance of radioactive cesium in domestic foods on the Japanese market, J Food Soc Jpn, 56(2):49-56, 2015.
[2] D. Rangaswamvi et al, Estimation of radiological dose from radon, thoron and their progeny levels in the dwellings of Shivamogga district, Proc IARP, 2018.
[3] S. Miyuki et al, New method for visualizing the dose rate distribution around the Fukushima Daiichi nuclear power plant using ANNs, Sci Rep 11, 1857, 2021.
[4] S. Huseyin, Gamma Spectroscopy by Artificial Neural Network Coupled with MCNP, PhD thesis, 2017.
[5] C. Lattner, "LLVM: An Infrastructure for Multi-Stage Optimization," M.S. thesis, Dept. of Computer Science, Univ. of Illinois, USA, 2002.
[6] C. Lattner, "LLVM: A Compilation Framework for Lifelong Program Analysis & Transformation," Ph.D. thesis, Dept. of Computer Science, Univ. of Illinois, USA, 2005.
[7] Apple, Mixing Swift and C++, Swift.org Blog, Junio 2023, disponible en: https://www.swift.org/documentation/cxx-interop/
[8] B. Jacob, Eigen C++ template library, sitio oficial, recuperado: Julio 2024, disponible en: http:// eigen.tuxfamily.org
[9] T. Burr & M. Hamada, Radio-Isotope Identification Algorithms for NaI γ Spectra, Algorith ms, 2(1): 339-360, 2009.
[10] B. Tom & H. Michael, Radio-isotope identification algorithms for NaI g spectra, Algorithms, 339-360, 2009.
[11] M. Alamaniotis et al, Fuzzy-Logic Radioisotope Identifier for Gamma Spectroscopy in Source Search, IEEE Trans. Nucl. Sci, 60(4): 3014-3024, 2013.
[12] B. David et al, Principal component analysis of gamma-ray spectra for radiation portal monitors, IEEE Trans. Nucl. Sci., 59(1):154-160, 2012.
[13] S. Galib et al, A comparative study of machine learning methods for automated identificati on of radioisotopes using NaI gamma-ray spectra, Nucl. Eng. Tech., 53: 4072-4079, 2021.
[14] S. Huseyin, Gamma Spectroscopy by Artificial Neural Network Coupled with MCNP, PhD thesis, 2017.
[15] A. Miltiadis et al, Application of Fireworks Algorithm in Gamma-Ray Spe ctrum Fitting for Radioisotope Identification, Int. J. Swarm Intel. Res. (IJSIR), 6(6):102-125, 2015.
[16] A. Miltiadis et al, Fuzzy-logic radioisotope identifier for gamma spectroscopy in source search, IEEE Trans. Nucl. Sci., 60(4):3014-3024, 2013.
[17] R. Runkle et al, Analysis of spectroscopic radiation portal monitor data using principal components analysis, IEEE Trans. Nucl. Sci., 53(3):1418-1423, 2016.
[18] Apple, A Vision for Embedded Swift, Swift GitHub repository, Junio 2024, disponible en: http://github.com/sw iftlang/swift-evolution/blob/main/visions/embedded-swift.md
[19] K. Mracek, Get Started with Embedded Swift on ARM and RISC-V Microcontrollers, Swift Blog, Abril 2024, disponible en: https://www.swift.org/blog/embedded-swift-examples
CITAR COMO:
Pacheco-González, A.; Rivera, M.; Chacón-Blanco, R.; Robledo-Vega, Isidro, "Identificación de Radioisotopos en Tiempo Real Aplicando Mínimos Cuadrados Usando C++ para una Arquitectura Embebida", Revista ELECTRO, Vol. 46, 2024, pp. 37-42.
VERSIÓN PDF