Physics informed neural networks for surrogate modeling of accidental scenarios in nuclear power plants
2
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Licensing the next-generation of nuclear reactor designs requires extensive use of Modeling and Simulation (M&S) to investigate system response to many operational conditions, identify possible accidental scenarios and predict their evolution to undesirable consequences that are to be prevented or mitigated via the deployment of adequate safety barriers. Deep Learning (DL) and Artificial Intelligence (AI) can support M&S computationally by providing surrogates of the complex multi-physics high-fidelity models used for design. However, DL and AI are, generally, low-fidelity ‘black-box’ models that do not assure any structure based on physical laws and constraints, and may, thus, lack interpretability and accuracy of the results. This poses limitations on their credibility and doubts about their adoption for the safety assessment and licensing of novel reactor designs. In this regard, Physics Informed Neural Networks (PINNs) are receiving growing attention for their ability to integrate fundamental physics laws and domain knowledge in the neural networks, thus assuring credible generalization capabilities and credible predictions. This paper presents the use of PINNs as surrogate models for accidental scenarios simulation in Nuclear Power Plants (NPPs). A case study of a Loss of Heat Sink (LOHS) accidental scenario in a Nuclear Battery (NB), a unique class of transportable, plug-and-play microreactors, is considered. A PINN is developed and compared with a Deep Neural Network (DNN). The results show the advantages of PINNs in providing accurate solutions, avoiding overfitting, underfitting and intrinsically ensuring physics-consistent results
Abstract
Licensing the next-generation of nuclear reactor designs requires extensive use of Modeling and Simulation (M&S) to investigate system response to many operational conditions, identify possible accidental scenarios and predict their evolution to undesirable consequences that are to be prevented or mitigated via the deployment of adequate safety barriers. Deep Learning (DL) and Artificial Intelligence (AI) can support M&S computationally by providing surrogates of the complex multi-physics high-fidelity models used for design. However, DL and AI are, generally, low-fidelity ‘black-box’ models that do not assure any structure based on physical laws and constraints, and may, thus, lack interpretability and accuracy of the results. This poses limitations on their credibility and doubts about their adoption for the safety assessment and licensing of novel reactor designs. In this regard, Physics Informed Neural Networks (PINNs) are receiving growing attention for their ability to integrate fundamental physics laws and domain knowledge in the neural networks, thus assuring credible generalization capabilities and credible predictions. This paper presents the use of PINNs as surrogate models for accidental scenarios simulation in Nuclear Power Plants (NPPs). A case study of a Loss of Heat Sink (LOHS) accidental scenario in a Nuclear Battery (NB), a unique class of transportable, plug-and-play microreactors, is considered. A PINN is developed and compared with a Deep Neural Network (DNN). The results show the advantages of PINNs in providing accurate solutions, avoiding overfitting, underfitting and intrinsically ensuring physics-consistent results
참고문헌 (45)
- Majdi I. Radaideh, "Surrogate modeling of advanced computer simulations using deep Gaussian processes", Reliability Engineering & System Safety, 195, pp.106731~ (2020), Elsevier BV[DOI]
- F. D’Auria, "The Best Estimate Plus Uncertainty (BEPU) approach in licensing of current nuclear reactors", Nuclear Engineering and Design, 248, pp.317~328 (2012), Elsevier BV[DOI]
- Xu Wu, "Inverse uncertainty quantification of reactor simulations under the Bayesian framework using surrogate models constructed by polynomial chaos expansion", Nuclear Engineering and Design, 313, pp.29~52 (2017), Elsevier BV[DOI]
- Rohit K. Tripathy, "Deep UQ: Learning deep neural network surrogate models for high dimensional uncertainty quantification", Journal of Computational Physics, 375, pp.565~588 (2018), Elsevier BV[DOI]
- Majdi I. Radaideh, "Analyzing nuclear reactor simulation data and uncertainty with the group method of data handling", Nuclear Engineering and Technology, 52(2), pp.287~295 (2020), 한국원자력학회
- Peter Benner, "A Survey of Projection-Based Model Reduction Methods for Parametric Dynamical Systems", SIAM Review, 57(4), pp.483~531 (2015), Society for Industrial & Applied Mathematics (SIAM)[DOI]
- S. Chaturantabut, "Nonlinear model reduction via discrete empirical interpolation", SIAM J. Sci. Comput, 32(5), pp.2737~2764 (2010)
- Clarence Worrell, "Machine learning of fire hazard model simulations for use in probabilistic safety assessments at nuclear power plants", Reliability Engineering & System Safety, 183, pp.128~142 (2019), Elsevier BV[DOI]
- Jorge E. Hurtado, "Filtered importance sampling with support vector margin: A powerful method for structural reliability analysis", Structural Safety, 29(1), pp.2~15 (2007), Elsevier BV[DOI]
- Giovanni Roma, "Passive safety systems analysis: A novel approach for inverse uncertainty quantification based on Stacked Sparse Autoencoders and Kriging metamodeling", Progress in Nuclear Energy, 148, pp.104209~ (2022), Elsevier BV[DOI]
- 윤석, "Thermal conductivity prediction model for compacted bentonites considering temperature variations", Nuclear Engineering and Technology, 53(10), pp.3359~3366 (2021), 한국원자력학회[DOI]
- Joseph P. Yurko, "Demonstration of Emulator-Based Bayesian Calibration of Safety Analysis Codes: Theory and Formulation", Science and Technology of Nuclear Installations, 2015, pp.1~17 (2015), Hindawi Limited[DOI]
- Bamidele Ebiwonjumi, "Uncertainty quantification of PWR spent fuel due to nuclear data and modeling parameters", Nuclear Engineering and Technology, 53(3), pp.715~731 (2021), 한국원자력학회[DOI]
- B. Sudret, "Computing derivative-based global sensitivity measures using polynomial chaos expansions", Reliability Engineering & System Safety, 134, pp.241~250 (2015), Elsevier BV[DOI]
- L. Puppo, "A Framework based on Finite Mixture Models and Adaptive Kriging for Characterizing Non-Smooth and Multimodal Failure Regions in a Nuclear Passive Safety System", Reliability Engineering & System Safety, 216, pp.107963~ (2021), Elsevier BV[DOI]
- Nicola Pedroni, "An Adaptive Metamodel-Based Subset Importance Sampling approach for the assessment of the functional failure probability of a thermal-hydraulic passive system", Applied Mathematical Modelling, 48, pp.269~288 (2017), Elsevier BV[DOI]
- Xueguan Song, "A radial basis function-based multi-fidelity surrogate model: exploring correlation between high-fidelity and low-fidelity models", Structural and Multidisciplinary Optimization, 60(3), pp.965~981 (2019), Springer Science and Business Media LLC[DOI]
- L. Conner, "Machine learned metamodeling of a computationally intensive accident simulation code", Int. Conf. Nucl. Eng. Proc., ICONE, 1, pp.1~6 (2021)
- Rabie Abu Saleem, "Application of deep neural networks for high-dimensional large BWR core neutronics", Nuclear Engineering and Technology, 52(12), pp.2709~2716 (2020), 한국원자력학회[DOI]
- Luning Sun, "Surrogate modeling for fluid flows based on physics-constrained deep learning without simulation data", Computer Methods in Applied Mechanics and Engineering, 361, pp.112732~ (2020), Elsevier BV[DOI]
- Abiodun Ayodeji, "Deep learning for safety assessment of nuclear power reactors: Reliability, explainability, and research opportunities", Progress in Nuclear Energy, 151, pp.104339~ (2022), Elsevier BV[DOI]
- Valentina Ciriello, "Polynomial chaos expansion for global sensitivity analysis applied to a model of radionuclide migration in a randomly heterogeneous aquifer", Stochastic Environmental Research and Risk Assessment, 27(4), pp.945~954 (2012), Springer Science and Business Media LLC[DOI]
- Forrest Shriver, "Prediction of Neutronics Parameters Within a Two-Dimensional Reflective PWR Assembly Using Deep Learning", Nuclear Science and Engineering, 195(6), pp.626~647 (2021), Informa UK Limited[DOI]
- M.I. Radaideh, "Modeling nuclear data uncertainties using deep neural networks", International Conference on Physics of Reactors:Transition to a Scalable Nuclear Future, PHYSOR 2020, pp.2583~2590 (2020)
- F. Cadini, "Improved metamodel-based importance sampling for the performance assessment of radioactive waste repositories", Reliability Engineering & System Safety, 134, pp.188~197 (2015), Elsevier BV[DOI]
- Dean Price, "Multiobjective optimization of nuclear microreactor reactivity control system operation with swarm and evolutionary algorithms", Nuclear Engineering and Design, 393, pp.111776~ (2022), Elsevier BV[DOI]
- Hyeonmin Kim, "Application of a Deep Learning Technique to the Development of a Fast Accident Scenario Identifier", IEEE Access, 8, pp.177363~177373 (2020), Institute of Electrical and Electronics Engineers (IEEE)[DOI]
- Seunghyoung Ryu, "Probabilistic deep learning model as a tool for supporting the fast simulation of a thermal–hydraulic code", Expert Systems with Applications, 200, pp.116966~ (2022), Elsevier BV[DOI]
- Majid Ebad Sichani, "Probabilistic seismic analysis of concrete dry cask structures", Structural Safety, 73, pp.87~98 (2018), Elsevier BV[DOI]
- Rijan Shrestha, "Inverse uncertainty quantification of input model parameters for thermal-hydraulics simulations using expectation–maximization under Bayesian framework", Journal of Applied Statistics, 43(6), pp.1011~1026 (2015), Informa UK Limited[DOI]
- Alejandro Barredo Arrieta, "Explainable Artificial Intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AI", Information Fusion, 58, pp.82~115 (2020), Elsevier BV[DOI]
- George Em Karniadakis, "Physics-informed machine learning", Nature Reviews Physics, 3(6), pp.422~440 (2021), Springer Science and Business Media LLC[DOI]
- M. Raissi, "Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations", Journal of Computational Physics, 378, pp.686~707 (2019), Elsevier BV[DOI]
- Shengze Cai, "Physics-Informed Neural Networks for Heat Transfer Problems", Journal of Heat Transfer, 143(6) (2021), ASME International[DOI]
- Enrico Schiassi, "Physics-informed neural networks for the point kinetics equations for nuclear reactor dynamics", Annals of Nuclear Energy, 167, pp.108833~ (2022), Elsevier BV[DOI]
- Ehsan Haghighat, "A physics-informed deep learning framework for inversion and surrogate modeling in solid mechanics", Computer Methods in Applied Mechanics and Engineering, 379, pp.113741~ (2021), Elsevier BV[DOI]
- Yutian Lu, "Physics-informed surrogate modeling for hydro-fracture geometry prediction based on deep learning", Energy, 253, pp.124139~ (2022), Elsevier BV[DOI]
- Jacopo Buongiorno, "Can Nuclear Batteries Be Economically Competitive in Large Markets?", Energies, 14(14), pp.4385~ (2021), MDPI AG[DOI]
- Federico Antonello, "A methodology to perform dynamic risk assessment using system theory and modeling and simulation: Application to nuclear batteries", Reliability Engineering & System Safety, 228, pp.108769~ (2022), Elsevier BV[DOI]
- Federico Antonello, "Insights in the safety analysis of an early microreactor design", Nuclear Engineering and Design, 404, pp.112203~ (2023), Elsevier BV[DOI]
- Enrico Zio, "Prognostics and Health Management (PHM): Where are we and where do we (need to) go in theory and practice", Reliability Engineering & System Safety, 218, pp.108119~ (2022), Elsevier BV[DOI]
- A. F . Henry, "The Application of Reactor Kinetics to the Analysis of Experiments", Nuclear Science and Engineering, 3(1), pp.52~70 (1958), Informa UK Limited[DOI]
- D. P. Kingma, "Adam: a method for stochastic optimization", 3rd Int. Conf. Learn. Represent. ICLR 2015 - Conf.Track Proc, pp.1~15 (2015)
- "Comparison of Non-linear Activation Functions for Deep Neural Networks on MNIST Classification Task, vol. 3" (2018)[URL]
- C. Cortes, "L 2 regularization for learning kernels", Twenty-Fifth Conference on Uncertainty in Artificial Intelligence, pp.109~116 (2004)