arXiv subjects

Jean-Philippe Argaud

Publications and source records attributed to Jean-Philippe Argaud.

15 recordsLinked to original sources

arXiv2025

Artificial Intelligence in Reactor Physics: Current Status and Future Prospects

Reactor physics is the study of neutron properties, focusing on using models to examine the interactions between neutrons and materials in nuclear reactors. Artificial intelligence (AI) has made significant contributions to reactor physics, e.g., in operational simulations, safety design, real-time monitoring, core management and maintenance. This paper presents a comprehensive review of AI approaches in reactor physics, especially considering the category of Machine Learning (ML), with the aim of describing the application scenarios, frontier topics, unsolved challenges and future research directions. From equation solving and state parameter prediction to nuclear industry applications, this paper provides a step-by-step overview of ML methods applied to steady-state, transient and combustion problems. Most literature works achieve industry-demanded models by enhancing the efficiency of deterministic methods or correcting uncertainty methods, which leads to successful applications. However, research on ML methods in reactor physics is somewhat fragmented, and the ability to generalize models needs to be strengthened. Progress is still possible, especially in addressing theoretical challenges and enhancing industrial applications such as building surrogate models and digital twins.

cs.LG↗

arXiv2024

Projection-based model order reduction for prestressed concrete with an application to the standard section of a nuclear containment building

We propose a projection-based model order reduction procedure for the ageing of large prestressed concrete structures. Our work is motivated by applications in the nuclear industry, particularly in the simulation of containment buildings. Such numerical simulations involve a multi-modeling approach: a three-dimensional nonlinear thermo-hydro-visco-elastic rheological model is used for concrete; and prestressing cables are described by a one-dimensional linear thermo-elastic behavior. A kinematic linkage is performed in order to connect the concrete nodes and the steel nodes: coincident points in each material are assumed to have the same displacement. We develop an adaptive algorithm based on a Proper Orthogonal Decomposition (POD) in time and greedy in parameter to build a reduced order model (ROM). The nonlinearity of the operator entails that the computational cost of the ROM assembly scales with the size of the high-fidelity model. We develop an hyper-reduction strategy based on empirical quadrature to bypass this computational bottleneck: our approach relies on the construction of a reduced mesh to speed up online assembly costs of the ROM. We provide numerical results for a standard section of a double-walled containment building using a qualified and broadly-used industrial grade finite element solver for structural mechanics (code$\_$aster).

math.NA↗

arXiv2023

Exact and general decoupled solutions of the LMC Multitask Gaussian Process model

The Linear Model of Co-regionalization (LMC) is a very general multitask gaussian process model for regression or classification. While its expressiveness and conceptual simplicity are appealing, naive implementations have cubic complexity in the product (number of datapoints $\times$ number of tasks), making approximations mandatory for most applications. However, recent work has shown that in some settings the latent processes of the model can be decoupled, leading to a complexity that is only linear in the number of said processes. We here extend these results, showing from the most general assumptions that the only condition necessary to an efficient exact computation of the LMC is a mild hypothesis on the noise model. We introduce a full parametrization of the resulting \emph{projected LMC} model, enabling its efficient optimization. The effectiveness of this approach is assessed through synthetic and real-data experiments, testing in particular the behavior of its underlying noise model restriction.\\ Overall, the projected LMC appears as a competitive and simpler alternative to state-of-the art multitask gaussian process models. It greatly facilitates some computations such as training data updates or leave-one-out cross-validation, and is more interpretable, for it gives access to its low-dimensional quantities and to their explicit relation with the full-dimensional data. These qualities could facilitate the adoption by various industries of entire classes of methodologies, notably multitask bayesian optimization.

cs.LG↗

arXiv2022

A projection-based reduced-order model for parametric quasi-static nonlinear mechanics using an open-source industrial code

We propose a projection-based model order reduction procedure for a general class of parametric quasi-static problems in nonlinear mechanics with internal variables. The methodology is integrated in the industrial finite element code code aster. Model order reduction aims to lower the computational cost of engineering studies that involve the simulation to a costly high-fidelity differential model for many different parameters, which correspond, for example to material properties or initial and boundary conditions. We develop an adaptive algorithm based on a POD-Greedy strategy, and we develop an hyper-reduction strategy based on an element-wise empirical quadrature in order to speed up the assembly costs of the reduced-order model by building an appropriate reduced mesh. We introduce a cost-efficient error indicator which relies on the reconstruction of the stress field by a Gappy-POD strategy. We present numerical results for a three-dimensional elastoplastic system in order to illustrate and validate the methodology.

math.NA↗

arXiv2021

Observation data compression for variational assimilation of dynamical systems

Accurate estimation of error covariances (both background and observation) is crucial for efficient observation compression approaches in data assimilation of large-scale dynamical problems. We propose a new combination of a covariance tuning algorithm with existing PCA-type data compression approaches, either observation- or information-based, with the aim of reducing the computational cost of real-time updating at each assimilation step. Relying on a local assumption of flow-independent error covariances, dynamical assimilation residuals are used to adjust the covariance in each assimilation window. The estimated covariances then contribute to better specify the principal components of either the observation dynamics or the state-observation sensitivity. The proposed approaches are first validated on a shallow water twin experiment with correlated and non-homogeneous observation error. Proper selection of flow-independent assimilation windows, together with sampling density for background error estimation, and sensitivity of the approaches to the observations error covariance knowledge, are also discussed and illustrated with various numerical tests and results. The method is then applied to a more challenging industrial hydrological model with real-world data and a non-linear transformation operator provided by an operational precipitation-flow simulation software.

math.NA↗

arXiv2020

A graph clustering approach to localization for adaptive covariance tuning in data assimilation based on state-observation mapping

An original graph clustering approach to efficient localization of error covariances is proposed within an ensemble-variational data assimilation framework. Here the localization term is very generic and refers to the idea of breaking up a global assimilation into subproblems. This unsupervised localization technique based on a linearizedstate-observation measure is general and does not rely on any prior information such as relevant spatial scales, empirical cut-off radius or homogeneity assumptions. It automatically segregates the state and observation variables in an optimal number of clusters (otherwise named as subspaces or communities), more amenable to scalable data assimilation.The application of this method does not require underlying block-diagonal structures of prior covariance matrices. In order to deal with inter-cluster connectivity, two alternative data adaptations are proposed. Once the localization is completed, an adaptive covariance diagnosis and tuning is performed within each cluster. Numerical tests show that this approach is less costly and more flexible than a global covariance tuning, and most often results in more accurate background and observations error covariances.

math.ST↗

arXiv2019

Background Error Covariance Iterative Updating with Invariant Observation Measures for Data Assimilation

In order to leverage the information embedded in the background state and observations, covariance matrices modelling is a pivotal point in data assimilation algorithms. These matrices are often estimated from an ensemble of observations or forecast differences. Nevertheless, for many industrial applications the modelling still remains empirical based on some form of expertise and physical constraints enforcement in the absence of historical observations or predictions. We have developed two novel robust adaptive assimilation methods named CUTE (Covariance Updating iTerativE) and PUB (Partially Updating BLUE). These two non-parametric methods are based on different optimization objectives, both capable of sequentially adapting background error covariance matrices in order to improve assimilation results under the assumption of a good knowledge of the observation error covariances. We have compared these two methods with the standard approach using a misspecified background matrix in a shallow water twin experiments framework with a linear observation operator. Numerical experiments have shown that the proposed methods bear a real advantage both in terms of posterior error correlation identification and assimilation accuracy.

stat.ME↗

arXiv2014

An Advanced Computational Scheme for the Optimization of 2D Radial Reflectors in Pressurized Water Reactors

This paper presents a computational scheme for the determination of equivalent 2D multi-group heterogeneous reflectors in a Pressurized Water Reactor (PWR). The proposed strategy is to define a full-core calculation consistent with a reference lattice code calculation such as the Method Of Characteristics (MOC) as implemented in APOLLO2 lattice code. The computational scheme presented here relies on the data assimilation module known as "Assimilation de données et Aide à l'Optimisation (ADAO)" of the SALOME platform developed at Électricité De France (EDF), coupled with the full-core code COCAGNE and with the lattice code APOLLO2. A first validation of the computational scheme is made using the OPTEX reflector model developed at École Polytechnique de Montréal (EPM). As a result, we obtain 2D multi-group, spatially heterogeneous 2D reflectors, using both diffusion or $\text{SP}_{\text{N}}$ operators. We observe important improvements of the power discrepancies distribution over the core when using reflectors computed with the proposed computational scheme, and the $\text{SP}_{\text{N}}$ operator enables additional improvements.

nucl-th↗

arXiv2014

Reconstruction by data assimilation of the inner temperature field from outer measurements in a thick pipe

The detailed knowledge of the inner skin temperature behavior is very important to evaluate and manage the aging of large pipes in cooling systems. We describe here a method to obtain this information as a function of outer skin temperature measurements, in space and time. This goal is achieved by mixing fine simulations and numerical methods such as impulse response and data assimilation. Demonstration is done on loads representing extreme transient stratification or thermal shocks. From a numerical point of view, the results of the reconstruction are outstanding, with a mean accuracy of the order of less than a half percent of the temperature values of the thermal transient.

physics.comp-ph↗

arXiv2013

Variational Assimilation for Xenon Dynamical Forecasts in Neutronic using Advanced Background Error Covariance Matrix

Data assimilation method consists in combining all available pieces of information about a system to obtain optimal estimates of initial states. The different sources of information are weighted according to their accuracy by the means of error covariance matrices. Our purpose here is to evaluate the efficiency of variational data assimilation for the xenon induced oscillations forecasts in nuclear cores. In this paper we focus on the comparison between 3DVAR schemes with optimised background error covariance matrix B and a 4DVAR scheme. Tests were made in twin experiments using a simulation code which implements a mono-dimensional coupled model of xenon dynamics, thermal, and thermal-hydraulic processes. We enlighten the very good efficiency of the 4DVAR scheme as well as good results with the 3DVAR one using a careful multivariate modelling of B.

physics.data-an↗

arXiv2011

Unified nuclear core activity map reconstruction using heterogeneous instruments with data assimilation

Evaluating the neutronic state of the whole nuclear core is a very important topic that have strong implication for nuclear core management and for security monitoring. The core state is evaluated using measurements. Usually, part of the measurements are used, and only one kind of instruments are taken into account. However, the core state evaluation should be more accurate when more measurements are collected in the core. But using information from heterogeneous sources is at glance a difficult task. This difficulty can be overcome by Data Assimilation techniques. Such a method allows to combine in a coherent framework the information coming from model and the one coming from various type of observations. Beyond the inner advantage to use heterogeneous instruments, this leads to obtain a significant increasing of the quality of neutronic global state reconstruction with respect to individual use of measures. In order to present this approach, we will introduce here the basic principles of data assimilation focusing on BLUE (Best Unbiased Linear Estimation). Then we will present the configuration of the method within the nuclear core problematic. Finally, we will present the results obtained on nuclear measurement coming from various instruments.

physics.data-an↗

arXiv2011

Best linear unbiased estimation of the nuclear masses

This paper presents methods to provide an optimal evaluation of the nuclear masses. The techniques used for this purpose come from data assimilation that allows combining, in an optimal and consistent way, information coming from experiment and from numerical model. Using all the available information, it leads to improve not only masses evaluations, but also to decrease uncertainties. Each newly evaluated mass value is associated with some accuracy that is sensibly reduced with respect to the values given in tables, especially in the case of the less well-known masses. In this paper, we first introduce a useful tool of data assimilation, the Best Linear Unbiased Estimation (BLUE). This BLUE method is applied to nuclear mass tables and some results of improvement are shown.

nucl-th↗

arXiv2011

Optimal design of measurement network for neutronic activity field reconstruction by data assimilation

Using data assimilation framework, to merge information from model and measurement, an optimal reconstruction of the neutronic activity field can be determined for a nuclear reactor core. In this paper, we focus on solving the inverse problem of determining an optimal repartition of the measuring instruments within the core, to get the best possible results from the data assimilation reconstruction procedure. The position optimisation is realised using Simulated Annealing algorithm, based on the Metropolis-Hastings one. Moreover, in order to address the optimisation computing challenge, algebraic improvements of data assimilation have been developed and are presented here.

physics.data-an↗

arXiv2010

Differential influence of instruments in nuclear core activity evaluation by data assimilation

The global activity fields of a nuclear core can be reconstructed using data assimilation. Data assimilation allows to combine measurements from instruments, and information from a model, to evaluate the best possible activity within the core. We present and apply a specific procedure which evaluates this influence by adding or removing instruments in a given measurement network (possibly empty). The study of various network configurations of instruments in the nuclear core establishes that influence of the instruments depends both on the independant instrumentation location and on the chosen network.

physics.data-an↗

arXiv2010

Robustness of nuclear core activity reconstruction by data assimilation

We apply a data assimilation techniques, inspired from meteorological applications, to perform an optimal reconstruction of the neutronic activity field in a nuclear core. Both measurements, and information coming from a numerical model, are used. We first study the robustness of the method when the amount of measured information decreases. We then study the influence of the nature of the instruments and their spatial repartition on the efficiency of the field reconstruction.

physics.data-an↗