Lithuanian Journal of Physics article / Lietuvos fizikos žurnalo straipsnis

APPLICATION OF ARTIFICIAL NEURAL NETWORK FOR THE IONIZING RADIATION PARTICLE IDENTIFICATION BY THE PLASTIC SCINTILLATION DETECTOR RESPONSE
Jevgenij Garankin and Artūras Plukis
Center for Physical Sciences and Technology, Savanorių 231, 02300 Vilnius, Lithuania
Email: jevgenij.garankin@ftmc.lt

Received 6 June 2022; revised 31 July 2022; accepted 13 September 2022

The separation of ionizing radiation particles is an important and challenging task, especially regarding neutrons and gamma rays. The separation of neutron and gamma radiation is necessary for safeguard purposes and control of nuclear reactions. Standard mathematical models of pulse analysis work well in the presence of large energy transfer (>1 MeV) from the particle to the detector. However, the quality of the separation decreases as the amount of transferred energy lessens, making it impossible to determine the exact type of particle at a sufficiently low-energy level.
In this work, an artificial neural network model was used to solve the problem of separation at low-energy levels. The supervised machine learning (ML) model was used to analyse pulses received from the polyethylene naphthalate (PEN) scintillation detector. Several data sets after the PEN exposure to neutron/gamma (combined ²³⁹PuBe and ²³⁸PuBe source), alpha (²³⁸Pu) and beta (⁹⁰Sr/⁹⁰Y) sources were used to train the models. The information obtained from the separation of neutrons and gamma particles was compared with the information obtained using standard pulses delayed fluorescence analysis methods. The obtained results showed that the model was able to separate particles in the fields of low- and high-energy transfer.
Keywords: radiation particle discrimination, ANN, machine learning, scintillation detectors

DIRBTINIO NEURONINIO TINKLO TAIKYMAS JONIZUOJANČIOSIOS SPINDULIUOTĖS DALELĖMS IDENTIFIKUOTI PAGAL PLASTIKINIO SCINTILIACINIO DETEKTORIAUS ATSAKĄ
Jevgenij Garankin, Artūras Plukis

Fizinių ir technologijos mokslų centras, Vilnius, Lietuva

Jonizuojančiosios spinduliuotės dalelių atskyrimas – svarbi ir sudėtinga užduotis, ypač neutronų ir gama spindulių atžvilgiu. Neutronų ir gama spinduliuotės atskyrimas yra būtinas radiacinės saugos tikslais ir branduolinių reakcijų kontrolei. Standartiniai matematiniai impulsų analizės modeliai gerai veikia, esant didelei dalelės energijos perdavai (>1 MeV). Atskyrimo kokybė prastėja mažėjant perduodamos energijos kiekiui, todėl neįmanoma nustatyti tikslaus dalelės tipo, esant gana žemam energijos perdavos lygiui.
Šiame darbe buvo panaudotas dirbtinio neuroninio tinklo modelis, sprendžiant atskyrimo, esant žemai energijos perdavai, problemą. ML (mašininio mokymosi) modelis buvo naudojamas analizuojant impulsus, gautus iš PEN (polietileno naftalato) scintiliacinio detektoriaus. Dirbtinio neuroninio tinklo modeliams mokyti buvo naudojami duomenų rinkiniai, gauti veikiant PEN detektorių alfa (²³⁸Pu šaltinis), beta (⁹⁰Sr/⁹⁰Y šaltinis) ir kombinuotu neutronų ir gama fotonų (²³⁹PuBe ir ²³⁸PuBe šaltinių mišinys) srautais. Neuroninio tinklo atsako duomenys buvo lyginami su standartiniais matematiniais jonizuojančiosios spinduliuotės atskyrimo būdais, kuriuose naudojamas greitosios ir uždelstosios fluorescencijos santykis. Gauti rezultatai parodė, kad modelis sugeba labai gerai atskirti daleles didelės energijos perdavos srityje, taip pat jis identifikuoja daleles ir mažos energijos perdavos srityje, tačiau nėra įrankių, leidžiančių patikrinti rezultatų patikimumą.

References / Nuorodos

[1] R. Kouzes, J. Ely, A. Lintereur, E. Siciliano, D. Stromswold, and M.L. Woodring, 3He Neutron Detector Pressure Effect and Comparison to Models, Technical Report (Pacific Northwest National Lab. (PNNL), Richland, WA, United States, 2010),
https://doi.org/10.2172/971884
[2] M. Grodzicka-Kobylka, T. Szczesniak, L. Swiderski, K. Brylew, M. Moszyński, J. Valiente-Dobón, P. Schotanus, K. Grodzicki, P. Mazerewicz, J. Szymanowski, and Z. Mianowska, Comparison of detectors with pulse shape discrimination capability for simultaneous detection of gamma-rays, slow and fast neutrons, Nucl. Instrum. Methods Phys. Res. A 1019, 165858 (2021),
https://doi.org/10.1016/j.nima.2021.165858
[3] G.H.V. Bertrand, M. Hamel, S. Normand, and F. Sguerra, Pulse shape discrimination between (fast or thermal) neutrons and gamma rays with plastic scintillators: State of the art, Nucl. Instrum. Methods Phys. Res. A 776, 114–128 (2015),
https://doi.org/10.1016/j.nima.2014.12.024
[4] P.N. Zhmurin, V.N. Lebedev, A.F. Adadurov, V.N. Pereymak, and Yu.A. Gurkalenko, Plastic scintillator for pulse shape neutrons and gamma quanta discrimination, Radiat. Meas. 62, 1–5 (2014),
https://doi.org/10.1016/j.radmeas.2013.12.013
[5] T. Yanagida, K. Watanabe, and Y. Fujimoto, Comparative study of neutron and gamma-ray pulse shape discrimination of anthracene, stilbene, and p-terphenyl, Nucl. Instrum. Methods Phys. Res. A 784, 111–114 (2015),
https://doi.org/10.1016/j.nima.2014.12.031
[6] P. Blanc, M. Hamel, L. Rocha, S. Normand, R. Pansu, F. Gorbert, and I. Lampre, in: Proceedings of 2013 3rd International Conference on Advancements in Nuclear Instrumentation, Measurement Methods and Their Applications (Marseille, France, 2013) pp. 1–9,
https://doi.org/10.1109/ANIMMA.2013.6727919
[7] W.G.J. Langeveld, M.J. King, J. Kwong, and D.T. Wakeford, Pulse shape discrimination algorithms, figures of merit, and gamma-rejection for liquid and solid scintillators, IEEE Trans. Nucl. Sci. 64(7), 1801–1809 (2017),
https://doi.org/10.1109/TNS.2017.2681654
[8] L. Su, L. Deng, W. Zhu, and S. Zhao, Statistical detection of weak pulse signal under chaotic noise based on Elman neural network, Wirel. Commun. Mob. Comput. 2020, 1–12 (2020),
https://doi.org/10.1155/2020/9653586
[9] T.S. Sanderson, C.D. Scott, M. Flaska, J.K. Polack, and S.A. Pozziet, in: Proceedings of 2012 IEEE Nuclear Science Symposium and Medical Imaging Conference Record (NSS/MIC) (2012) pp. 199–202,
https://doi.org/10.1109/NSSMIC.2012.6551092
[10] J. Griffiths, S. Kleinegesse, D. Saunders, R. Taylor, and A. Vacheret, Pulse shape discrimination and exploration of scintillation signals using convolutional neural network, Mach. Learn. Sci. Technol. 1(4), 045022 (2020),
https://doi.org/10.1088/2632-2153/abb781
[11] J. Garankin, A. Plukis, R. Plukienė, E. Lagzdina, and V. Remeikis, Identification of particles of ionizing radiation by the analysis of fluorescence pulse form of the thin pen film scintillator, IEEE Trans. Nucl. Sci. 65(2), 739–743 (2018),
https://doi.org/10.1109/TNS.2017.2785683
[12] Y. Efremenko, L. Fajt, M. Febbraro, F. Fischer, C. Hayward, R. Hodák, T. Kraetzschmar, B. Majorovits, D. Muenstermann, E. Öz, et al., Use of polyethylene naphthalate as a self-vetoing structural material, JINST 14, P07006 (2019),
https://arxiv.org/abs/1901.03579v2,
https://doi.org/10.1088/1748-0221/14/07/P07006
[13] F.V. Ordenes and R. Silipo, Machine learning for marketing on the KNIME Hub: The development of a live repository for marketing applications, J. Bus. Res. 137, 393–410 (2021),
https://doi.org/10.1016/j.jbusres.2021.08.036