Stefano Riva
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Stefano Riva has a total of 36 co-authors. Here are the top co-authors by number of shared publications:
- Antonio Cammi
Politecnico di Milano
15 shared publications - Carolina Introini
Politecnico di Milano
14 shared publications Carolina Introini
4 shared publications- Stefano Lorenzi
Politecnico di Milano
4 shared publications Antonio Cammi
3 shared publicationsJ. Nathan Kutz
2 shared publications
Latest Publications
- 11 November 2026
- Published by Elsevier in Annals of Nuclear Energy
- Vol. 243
Shallow decoding strategies for reduced order modelling and multi-fidelity applied to Monte Carlo neutronics calculations
- 11 November 2026
- Published by Elsevier in Annals of Nuclear Energy
- Vol. 243
Shallow decoding strategies for reduced order modelling and multi-fidelity applied to Monte Carlo neutronics calculations
- Open Access
- 3 November 2026
- Published by Taylor & Francis in Nuclear Science and Engineering
- p. 1-18
A Perturbation Theory Based Analysis of Reactivity Feedbacks During Transients in Molten Salt Reactors
- 3 November 2026
- Published by Taylor & Francis in Nuclear Science and Engineering
- p. 1-18
Open AccessA Perturbation Theory Based Analysis of Reactivity Feedbacks During Transients in Molten Salt Reactors
- Open Access
- 29 September 2026
- Published by Elsevier in Chemical Engineering Science
- Vol. 339
Towards efficient parametric state estimation with shallow recurrent decoder networks: Application to circulating fuel reactors
- 29 September 2026
- Published by Elsevier in Chemical Engineering Science
- Vol. 339
Open AccessTowards efficient parametric state estimation with shallow recurrent decoder networks: Application to circulating fuel reactors
- Open Access
- 2 September 2026
- Published by Elsevier in Chemical Engineering Science
- Vol. 338
Constrained sensing and reliable state estimation with shallow recurrent decoders on a TRIGA Mark II reactor
- 2 September 2026
- Published by Elsevier in Chemical Engineering Science
- Vol. 338
Open AccessConstrained sensing and reliable state estimation with shallow recurrent decoders on a TRIGA Mark II reactor
- Open Access
- 29 July 2026
- Published by Elsevier in Nuclear Engineering and Technology
- Vol. 58 (12)
Adaptive parametric reconstruction with the generalised empirical interpolation method on the DYNASTY loop with limited instrumentation
- 29 July 2026
- Published by Elsevier in Nuclear Engineering and Technology
- Vol. 58 (12)
Open AccessAdaptive parametric reconstruction with the generalised empirical interpolation method on the DYNASTY loop with limited instrumentation
- Open Access
- 27 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Plasma Science
- Vol. 54 (6) , 2704-2709
Parametric Dynamic Mode Decomposition for Nuclear Fusion: Applications to Magnetohydrodynamics
Magnetohydrodynamics (MHD) investigates the dynamics of electrically conducting fluids interacting with magnetic fields. This theory is widely utilized for magnetic confinement fusion problems since both thermonuclear plasmas and some of the relevant operating fluids flowing in the blanket can be described as conducting fluids subjected to magnetic fields. The intrinsically multiphysical nature of MHD problems leads to complex and computationally expensive models, making them unsuited for multiquery applications such as parameter optimization and sensitivity analysis, as well as for online monitoring and control. Reduced-order modeling (ROM) offers a promising path to reduce computational costs while preserving accuracy; however, its application for MHD problems is still in its preliminary stages. This study continues the investigation started by the authors on ROM techniques for MHD problems by considering a parametric version of the dynamic mode decomposition (DMD), aiming to retrieve a general linear representation of the parametric time-series MHD data. In particular, a Kalman filter (KF)-enhanced version of the technique is tested on a lead–lithium MHD flow subjected to a magnetic field of various intensities. The integration of KF in the reduced model approach aims at improving the accuracy of the reduced model by corrections based on data assimilated from sensors inside the geometry. Results demonstrate the reduced model potential to significantly reduce computational effort while maintaining acceptable accuracy, especially if coupled with the KF, making it suitable for fusion-relevant MHD applications.
- 27 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Plasma Science
- Vol. 54 (6) , 2704-2709
Open AccessParametric Dynamic Mode Decomposition for Nuclear Fusion: Applications to Magnetohydrodynamics
Magnetohydrodynamics (MHD) investigates the dynamics of electrically conducting fluids interacting with magnetic fields. This theory is widely utilized for magnetic confinement fusion problems since both thermonuclear plasmas and some of the relevant operating fluids flowing in the blanket can be described as conducting fluids subjected to magnetic fields. The intrinsically multiphysical nature of MHD problems leads to complex and computationally expensive models, making them unsuited for multiquery applications such as parameter optimization and sensitivity analysis, as well as for online monitoring and control. Reduced-order modeling (ROM) offers a promising path to reduce computational costs while preserving accuracy; however, its application for MHD problems is still in its preliminary stages. This study continues the investigation started by the authors on ROM techniques for MHD problems by considering a parametric version of the dynamic mode decomposition (DMD), aiming to retrieve a general linear representation of the parametric time-series MHD data. In particular, a Kalman filter (KF)-enhanced version of the technique is tested on a lead–lithium MHD flow subjected to a magnetic field of various intensities. The integration of KF in the reduced model approach aims at improving the accuracy of the reduced model by corrections based on data assimilated from sensors inside the geometry. Results demonstrate the reduced model potential to significantly reduce computational effort while maintaining acceptable accuracy, especially if coupled with the KF, making it suitable for fusion-relevant MHD applications.
- Open Access
- 16 January 2026
- Published by The Open Journal in Journal of Open Source Software
- Vol. 11 (117) , 6950
pyforce: Python Framework for data-driven model Order Reduction of multi-physiCs problEms
Pyforce (Python Framework for data-driven model Order Reduction of multi-physiCs problEms) is a Python library implementing Data-Driven Reduced Order Modelling (DDROM) techniques (Riva et al., 2024) for applications to multi-physics problems, mainly in the nuclear engineering world. These techniques have been implemented upon the dolfinx package (Baratta et al., 2023) (currently v0.6.0), part of the FEniCSx project, to handle mesh generation, integral calculation and functions storage. The package is part of the ROSE (Reduced Order modelling with data-driven techniques for multi-phySics problEms) framework, which is one of the main research topics investigated at the ERMETE-Lab: in particular, the focus of the research activities is on mathematical algorithms aimed at reducing the complexity of multi-physics models with a focus on nuclear reactor applications, searching for optimal sensor positions and integrating experimental data to improve the knowledge on the physical systems.
- 16 January 2026
- Published by The Open Journal in Journal of Open Source Software
- Vol. 11 (117) , 6950
Open Accesspyforce: Python Framework for data-driven model Order Reduction of multi-physiCs problEms
Pyforce (Python Framework for data-driven model Order Reduction of multi-physiCs problEms) is a Python library implementing Data-Driven Reduced Order Modelling (DDROM) techniques (Riva et al., 2024) for applications to multi-physics problems, mainly in the nuclear engineering world. These techniques have been implemented upon the dolfinx package (Baratta et al., 2023) (currently v0.6.0), part of the FEniCSx project, to handle mesh generation, integral calculation and functions storage. The package is part of the ROSE (Reduced Order modelling with data-driven techniques for multi-phySics problEms) framework, which is one of the main research topics investigated at the ERMETE-Lab: in particular, the focus of the research activities is on mathematical algorithms aimed at reducing the complexity of multi-physics models with a focus on nuclear reactor applications, searching for optimal sensor positions and integrating experimental data to improve the knowledge on the physical systems.
- 9 January 2026
- Published by Taylor & Francis in Nuclear Technology
- Vol. 212 (6) , 1454-1478
A Comparative Study of Parametric Dynamic Mode Decomposition Algorithms on Thermal-Hydraulics Case Studies
- 9 January 2026
- Published by Taylor & Francis in Nuclear Technology
- Vol. 212 (6) , 1454-1478
A Comparative Study of Parametric Dynamic Mode Decomposition Algorithms on Thermal-Hydraulics Case Studies
- Open Access
- 16 September 2025
- Published by EDP Sciences in EPJ Nuclear Sciences & Technologies
- Vol. 11, 55
Data-driven reduced order modelling with malfunctioning sensors recovery applied to the Molten Salt Reactor case
This work presents the use of two Data-Driven Reduced Order Modelling techniques in predicting the transient response of a Molten Salt Fast Reactor when one or more sensors fail and, thus, provide wrong information; Supervised Machine Learning techniques are used to compensate for the failed sensors. Data-Driven Reduced Order Modelling integrate the physical knowledge contained in high-fidelity mathematical models with that coming from data measured on the actual system. This enables refining and updating the mathematical model, and address the challenges related to local-only observations, allowing for global state estimation. These methods are of interest when both sources of information are present, albeit incomplete, as is the case of the Molten Salt Fast Reactor. In these designs, typically operating in the fast neutron spectrum, the fuel is liquid, and no solid structures are foreseen in the core, thus making sensing and monitoring of safety-critical parameters and quantities quite challenging. Additionally, most literature studies on Data-Driven Reduced Order Modelling take the experimental observations as (noisy) ground-truth: very few works consider the case in which sensor fail or malfunction, and how this affect the state estimation.
- 16 September 2025
- Published by EDP Sciences in EPJ Nuclear Sciences & Technologies
- Vol. 11, 55
Open AccessData-driven reduced order modelling with malfunctioning sensors recovery applied to the Molten Salt Reactor case
This work presents the use of two Data-Driven Reduced Order Modelling techniques in predicting the transient response of a Molten Salt Fast Reactor when one or more sensors fail and, thus, provide wrong information; Supervised Machine Learning techniques are used to compensate for the failed sensors. Data-Driven Reduced Order Modelling integrate the physical knowledge contained in high-fidelity mathematical models with that coming from data measured on the actual system. This enables refining and updating the mathematical model, and address the challenges related to local-only observations, allowing for global state estimation. These methods are of interest when both sources of information are present, albeit incomplete, as is the case of the Molten Salt Fast Reactor. In these designs, typically operating in the fast neutron spectrum, the fuel is liquid, and no solid structures are foreseen in the core, thus making sensing and monitoring of safety-critical parameters and quantities quite challenging. Additionally, most literature studies on Data-Driven Reduced Order Modelling take the experimental observations as (noisy) ground-truth: very few works consider the case in which sensor fail or malfunction, and how this affect the state estimation.
- 28 August 2025
- Published by Taylor & Francis in Nuclear Science and Engineering
- Vol. 200 (sup1)
Real-Time State Estimation of Neutron Flux in Molten Salt Fast Reactors from Out-Core Sparse Measurements
- 28 August 2025
- Published by Taylor & Francis in Nuclear Science and Engineering
- Vol. 200 (sup1)
Real-Time State Estimation of Neutron Flux in Molten Salt Fast Reactors from Out-Core Sparse Measurements