Application of one-dimensional convolutional neural networks for identifying weak radioactive signals in dynamic monitoring systems

Машкіна, Ірина Вікторівна and Носенко, Тетяна Іванівна and Мельник, Ірина Юріївна (2025) Application of one-dimensional convolutional neural networks for identifying weak radioactive signals in dynamic monitoring systems Кібербезпека: освіта, наука, техніка (3). pp. 780-791. ISSN 2663-4023

Text 056_3_31_781-791-1-11.pdf Download (1MB)

Abstract

The work is devoted to solving the urgent problem of automated detection of low-activity moving sources of ionising radiation in dynamic radiation monitoring systems . The shortcomings of classical threshold signal processing algorithms which are characterised by low sensitivity and a high level of false alarms in conditions of low signal-to-noise ratio, are analysed. A new method for identifying radiation anomalies based on a one-dimensional convolutional neural network (1D CNN) is proposed. A mathematical model and software algorithm for generating synthetic time series of a scintillation detector have been developed, taking into account the stochastic nature of radiation registration (Poisson statistics), the geometry of vehicle passage, and shielding effects . The developed neural network model was trained and tested on generated data sets. The effectiveness of the proposed approach has been experimentally confirmed : the accuracy of signal classification from sources of average activity was 94.2%, and for extremely weak signals that are not visually distinguishable against the background noise, it was 83.5%. The results obtained indicate the promise of applying deep learning methods to improve the reliability of radiation safety systems.

Item Type:	Article
Uncontrolled Keywords:	radiation monitoring; scintillation detector; convolutional neural network ; 1D CNN; deep learning; signal processing; moving radiation sources
Subjects:	Статті у періодичних виданнях > Фахові (входять до переліку фахових, затверджений МОН)
Divisions:	Факультет інформаційних технологій та математики > Кафедра математики і фізики
Depositing User:	доцент Ірина Юріївна Мельник
Date Deposited:	25 Dec 2025 10:10
Last Modified:	25 Dec 2025 10:10
URI:	https://elibrary.kubg.edu.ua/id/eprint/55217

Actions (login required)

View Item