AI-driven real-time control of radiological contamination and personnel dose in nuclear transient scenarios

IF 2.2 3区环境科学与生态学 Q3 ENVIRONMENTAL SCIENCES

Journal of environmental radioactivity Pub Date : 2026-01-01 Epub Date: 2025-10-27 DOI:10.1016/j.jenvrad.2025.107838

Behzad Khosrowpour, Vahid Jamshidi

{"title":"AI-driven real-time control of radiological contamination and personnel dose in nuclear transient scenarios","authors":"Behzad Khosrowpour, Vahid Jamshidi","doi":"10.1016/j.jenvrad.2025.107838","DOIUrl":null,"url":null,"abstract":"<div><div>This study presents an artificial intelligence (AI) system designed to optimize emergency responses during localized nuclear incidents involving containment breaches and gas filtration failures. The system integrates a feed-forward neural network (ANN) with optimization algorithms, including fmincon and ParticleSwarm Optimization (PSO), to manage critical parameters such as ventilation speed, activation timing, and personnel evacuation. A synthetic dataset of 3600 samples, simulating the dispersion of gamma and beta radiation, was generated using MATLAB. The ANN achieved a coefficient of determination (R<sup>2</sup>) of 0.9876 and a root mean square error (RMSE) of 0.3421. The AI system reduced radioactive concentration by 15 % and cumulative effective dose by approximately 30 % (1.8 mSv vs. 2.5 mSv) compared to a baseline scenario (fixed ventilation at 2 m/s, evacuation at 30 min), representing the minimum operational ventilation speed per Table 3, chosen to represent a conservative, non-optimized response typical of manual procedures (IAEA, 2018). Personnel exposure was maintained at 1.8 mSv over 30 min, well below the IAEA occupational limit of 20 mSv/year (annual context only), with decision cycles completed in under 3 s. This framework demonstrates the feasibility of AI-driven real-time nuclear emergency management, offering improved decision-making over traditional methods while ensuring compliance with IAEA GSR Part 3 and Part 7 standards.</div></div>","PeriodicalId":15667,"journal":{"name":"Journal of environmental radioactivity","volume":"291 ","pages":"Article 107838"},"PeriodicalIF":2.2000,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of environmental radioactivity","FirstCategoryId":"93","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0265931X25002255","RegionNum":3,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2025/10/27 0:00:00","PubModel":"Epub","JCR":"Q3","JCRName":"ENVIRONMENTAL SCIENCES","Score":null,"Total":0}

引用次数: 0

Abstract

This study presents an artificial intelligence (AI) system designed to optimize emergency responses during localized nuclear incidents involving containment breaches and gas filtration failures. The system integrates a feed-forward neural network (ANN) with optimization algorithms, including fmincon and ParticleSwarm Optimization (PSO), to manage critical parameters such as ventilation speed, activation timing, and personnel evacuation. A synthetic dataset of 3600 samples, simulating the dispersion of gamma and beta radiation, was generated using MATLAB. The ANN achieved a coefficient of determination (R²) of 0.9876 and a root mean square error (RMSE) of 0.3421. The AI system reduced radioactive concentration by 15 % and cumulative effective dose by approximately 30 % (1.8 mSv vs. 2.5 mSv) compared to a baseline scenario (fixed ventilation at 2 m/s, evacuation at 30 min), representing the minimum operational ventilation speed per Table 3, chosen to represent a conservative, non-optimized response typical of manual procedures (IAEA, 2018). Personnel exposure was maintained at 1.8 mSv over 30 min, well below the IAEA occupational limit of 20 mSv/year (annual context only), with decision cycles completed in under 3 s. This framework demonstrates the feasibility of AI-driven real-time nuclear emergency management, offering improved decision-making over traditional methods while ensuring compliance with IAEA GSR Part 3 and Part 7 standards.

查看原文本刊更多论文

核瞬态情景下人工智能驱动的辐射污染和人员剂量实时控制。

本研究提出了一种人工智能（AI）系统，旨在优化涉及安全壳泄漏和气体过滤故障的局部核事件期间的应急响应。该系统集成了前馈神经网络（ANN）和优化算法，包括fmincon和ParticleSwarm optimization (PSO)，以管理通风速度、激活时间和人员疏散等关键参数。利用MATLAB生成了3600个样本的合成数据集，模拟了γ和β辐射的色散。人工神经网络的决定系数（R2）为0.9876，均方根误差（RMSE）为0.3421。与基线情景（固定通风速度为2米/秒，疏散时间为30分钟）相比，人工智能系统将放射性浓度降低了15%，累计有效剂量降低了约30%（1.8毫西弗vs. 2.5毫西弗），代表了表3中最小的操作通风速度，被选为代表手动程序的保守、非优化响应（IAEA, 2018）。人员暴露在30分钟内保持在1.8毫西弗，远低于原子能机构每年20毫西弗的职业限值（仅限年度情况），决策周期在30秒内完成。该框架证明了人工智能驱动的实时核应急管理的可行性，在确保符合原子能机构GSR第3部分和第7部分标准的同时，改进了传统方法的决策。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Journal of environmental radioactivity 环境科学-环境科学

CiteScore

4.70

自引率

13.00%

发文量

209

审稿时长

73 days

期刊介绍： The Journal of Environmental Radioactivity provides a coherent international forum for publication of original research or review papers on any aspect of the occurrence of radioactivity in natural systems. Relevant subject areas range from applications of environmental radionuclides as mechanistic or timescale tracers of natural processes to assessments of the radioecological or radiological effects of ambient radioactivity. Papers deal with naturally occurring nuclides or with those created and released by man through nuclear weapons manufacture and testing, energy production, fuel-cycle technology, etc. Reports on radioactivity in the oceans, sediments, rivers, lakes, groundwaters, soils, atmosphere and all divisions of the biosphere are welcomed, but these should not simply be of a monitoring nature unless the data are particularly innovative.