Artificial Intelligence, Meta-Modelling, Digital Twins and Advanced Simulation for the Safety Analysis of Nuclear Systems

Dear Colleagues,

The increasing adoption of Best Estimate Plus Uncertainty (BEPU) methodology has garnered significant attention within the international nuclear science and engineering community for its ability to provide a more realistic and comprehensive assessment of safety margins in nuclear systems. Within the BEPU paradigm, system responses under a variety of uncertain conditions are typically explored through the use of high-fidelity best estimate (BE) computational codes. These simulations play a pivotal role in supporting optimal design decisions, safety assessments, and the formulation of accident prevention and mitigation strategies. Nevertheless, several intrinsic challenges hinder the efficient use of such BE models. Specifically, BE codes are often

• Computationally intensive, requiring considerable resources and time for a single simulation;
• High-dimensional, involving large sets of input parameters and output responses;
• Opaque or black-box in nature, where the underlying mathematical formulation is not explicitly accessible and highly nonlinear;
• Dynamical, capturing the evolution of physical phenomena over time;
• Subject to significant uncertainties, frequently stemming from limited or imprecise input data.

To address these challenges, the integration of Artificial Intelligence (AI), meta-modeling techniques, Digital Twin technologies, and advanced simulation methods has emerged as a promising pathway for enhancing the efficiency and robustness of safety evaluations in nuclear systems. These approaches offer novel opportunities for accelerating simulation workflows, thereby enabling real-time system monitoring and prediction and improving the handling of uncertainties.

This Special Issue seeks to bring together contributions from leading researchers, engineers, and practitioners in the fields of nuclear safety, computational science, and intelligent systems. The objective is to showcase recent methodological advancements, practical implementations, and theoretical insights related to the application of AI, surrogate modeling, and advanced simulation techniques within the BEPU framework. This Special Issue welcomes both theoretical and applied research that addresses the analysis, optimization, and decision-making processes in nuclear safety assessment through data-driven and hybrid approaches.

Topics of interest include, but are not limited to, the following:

• Sensitivity analysis: methods, frameworks and applications in relation to nuclear safety;
• Forward uncertainty quantification in high-fidelity BEPU analyses;
• Inverse uncertainty quantification, parameter calibration, and model updating;
• Surrogate modeling and meta-modeling for computational acceleration;
• AI and ML for nuclear safety evaluation;
• Digital Twins for predictive monitoring and risk-informed decision support;
• Characterization of failure domains and rare-event analysis;
• Quantification of safety margins under epistemic and aleatory uncertainties;
• Hybrid physics-informed and data-driven modeling approaches;
• Applications of reduced-order modeling and emulation techniques in nuclear thermal-hydraulics.

When submitting your manuscript, please indicate that the manuscript is targeted for this specific Special Issue. For submission instructions, see Instructions for Authors.

Prof. Dr. Enrico Zio
Guest Editor

Dr. Ibrahim Ahmed
Co-Guest Editor

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Journal of Nuclear Engineering is an international peer-reviewed open access quarterly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 1200 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

• nuclear systems
• safety analysis
• meta-modeling
• digital twin
• artificial intelligence
• machine learning