A robust surrogate model of a solid oxide cell based on an adaptive polynomial approximation method

Abstract Surrogate models that predict the behaviors of solid oxide cells (SOCs) accurately at low computational cost are crucial to the control and optimization of SOC plants. Lumped physical models of SOCs, while widely used in such applications, lack accuracy because of neglected physical details. Data-driven models are the other options of surrogate models, which are proved to be more accurate because these models are identified directly from experiments or numerical simulations. However, due to the time cost of experiments and numerical simulations of SOCs, it is hoped that data-driven models can be constructed from a minimum amount of data. Also, the trained data-driven models should be robust, in other words, insensitive to the data set as well as the initial settings. These requirements are hard to be achieved by existing data-driven models of SOCs, such as lookup tables and artificial neural networks (ANNs). Aiming to preserve robustness and reduce the required amount of data, this paper introduces an adaptive polynomial approximation (APA) method, which is derived from the latest findings of numerical computation science, to the surrogate modeling of SOCs. The obtained models by the APA method are validated by both experiments and simulations. By analyzing the models, the coupling relationship among operating parameters of SOCs is revealed. The physical interpretability makes the APA method distinctive from common data-driven modeling methods. Performance comparison shows that the APA method is more accurate and robust than the existing ones with similar sampling costs. Additionally, the APA method can control the accuracy of the model by setting an error criterion in the algorithm iteration, endowing the APA method with an error control ability as per different accuracy requirements for SOC modeling.

[1]  Sanjay,et al.  Computational analysis of IR-SOFC: Transient, thermal stress, carbon deposition and flow dependency , 2016 .

[2]  Ellen Ivers-Tiffée,et al.  Electrochemical model for SOFC and SOEC mode predicting performance and efficiency , 2014 .

[3]  Xiongwen Zhang,et al.  Numerical investigations for a solid oxide electrolyte cell stack , 2019, International Journal of Hydrogen Energy.

[4]  Jaime Arriagada,et al.  Artificial neural network simulator for SOFC performance prediction , 2002 .

[5]  F. Jabbari,et al.  Investigation of thermal control for different SOFC flow geometries , 2016 .

[6]  Cosmin Safta,et al.  Toward using surrogates to accelerate solution of stochastic electricity grid operations problems , 2014, 2014 North American Power Symposium (NAPS).

[7]  Yixiang Shi,et al.  A robust flat-chip solid oxide fuel cell coupled with catalytic partial oxidation of methane , 2018, Journal of Power Sources.

[8]  Marco Sorrentino,et al.  A Review on solid oxide fuel cell models , 2011 .

[9]  Ahmed Fathy,et al.  Maximizing SOFC performance through optimal parameters identification by modern optimization algorithms , 2019, Renewable Energy.

[10]  Carl M. Stoots,et al.  3D CFD model of a multi-cell high-temperature electrolysis stack , 2007 .

[11]  Dongbin Xiu,et al.  The Wiener-Askey Polynomial Chaos for Stochastic Differential Equations , 2002, SIAM J. Sci. Comput..

[12]  Jarosław Milewski,et al.  Modelling the SOFC behaviours by artificial neural network , 2009 .

[13]  Andrei Kulikovsky.  A simple equation for temperature gradient in a planar SOFC stack , 2010 .

[14]  Gianni Bidini,et al.  Comparison of support vector regression and random forest algorithms for estimating the SOFC output voltage by considering hydrogen flow rates , 2020 .

[15]  Roland Pulch,et al.  Model order reduction for random nonlinear dynamical systems and low-dimensional representations for their quantities of interest , 2017, Math. Comput. Simul..

[16]  Jarosław Milewski,et al.  Modeling electrical behavior of solid oxide electrolyzer cells by using artificial neural network , 2015 .

[17]  Xin-Jian Zhu,et al.  Predictive control of SOFC based on a GA-RBF neural network model , 2008 .

[18]  Ionel M. Navon,et al.  Non-intrusive reduced order modelling of the Navier-Stokes equations , 2015 .

[19]  Jin Lin,et al.  Maximum Production Point Tracking of a High-Temperature Power-to-Gas System: A Dynamic-Model-Based Study , 2018, IEEE Transactions on Sustainable Energy.

[20]  Guangyi Cao,et al.  Nonlinear modeling of a SOFC stack based on a least squares support vector machine , 2006 .

[21]  O. Deutschmann,et al.  Dynamic behavior and control strategy study of CO2/H2O co-electrolysis in solid oxide electrolysis cells , 2019, Journal of Power Sources.

[22]  Xin-Jian Zhu,et al.  Short communication Modeling a SOFC stack based on GA-RBF neural networks identification , 2007 .

[23]  Danhui Gao,et al.  Fault tolerance control of SOFC systems based on nonlinear model predictive control , 2017 .

[24]  Hazlie Mokhlis,et al.  Optimization strategies for Solid Oxide Fuel Cell (SOFC) application: A literature survey , 2017 .

[25]  O. Deutschmann,et al.  Hierarchical modeling of solid oxide cells and stacks producing syngas via H2O/CO2 Co-electrolysis for industrial applications , 2018, Applied Energy.

[26]  Michael S. Eldred,et al.  Sparse Pseudospectral Approximation Method , 2011, 1109.2936.

[27]  Francisco Jurado.  Power supply quality improvement with a SOFC plant by neural-network-based control , 2003 .

[28]  Qiang Hu,et al.  Optimization of Hydrogen Yield of a High-Temperature Electrolysis System with Coordinated Temperature and Feed Factors at Various Loading Conditions: A Model-Based Study , 2018, Applied Energy.

[29]  Mohsen Hamedi,et al.  Modeling and Optimization of Anode‐Supported Solid Oxide Fuel Cells on Cell Parameters via Artificial Neural Network and Genetic Algorithm , 2012 .

[30]  Erdogan Guk,et al.  Parameters and their impacts on the temperature distribution and thermal gradient of solid oxide fuel cell , 2019, Applied Energy.

[31]  Alberto Traverso,et al.  Operating strategies to minimize degradation in fuel cell gas turbine hybrids , 2017 .

[32]  S. Jensen,et al.  Eliminating degradation in solid oxide electrochemical cells by reversible operation. , 2015, Nature Materials.

[33]  L. Barelli,et al.  Part load operation of SOFC/GT hybrid systems: Stationary analysis , 2012 .

[34]  Zhengkai Tu,et al.  Modelling of solid oxide electrolyser cell using extreme learning machine , 2017 .

[35]  F. J. Asensio,et al.  Model for optimal management of the cooling system of a fuel cell-based combined heat and power system for developing optimization control strategies , 2018 .

[36]  J. O’Brien,et al.  High-temperature electrolysis for large-scale hydrogen and syngas production from nuclear energy: summary of system simulation and economic analyses , 2010 .

[37]  N. Brandon,et al.  Hydrogen production through steam electrolysis: Control strategies for a cathode-supported intermediate temperature solid oxide electrolysis cell , 2008 .

[38]  Yixiang Shi,et al.  Flat-chip flame fuel cell operated on a catalytically enhanced porous media combustor , 2019, Energy Conversion and Management.

[39]  Carl M. Stoots,et al.  Computational Fluid Dynamics Model of a Planar Solid-Oxide Electrolysis Cell for Hydrogen Production from Nuclear Energy , 2007 .

[40]  Miriam Kemm,et al.  Steady state and transient thermal stress analysis in planar solid oxide fuel cells , 2005 .

[41]  Cheng Bao,et al.  Macroscopic modeling of solid oxide fuel cell (SOFC) and model-based control of SOFC and gas turbine hybrid system , 2018 .

[42]  Jiong Shen,et al.  Constrained model predictive control of a solid oxide fuel cell based on genetic optimization , 2011 .

[43]  G. Alexits.  Approximation theory , 1983 .

[44]  Ludger Blum,et al.  Bypassing renewable variability with a reversible solid oxide cell plant , 2018 .

[45]  Daejong Kim,et al.  Computational model to predict thermal dynamics of planar solid oxide fuel cell stack during start-up process , 2010 .

[46]  Jonathan Deseure,et al.  Solid oxide electrolysis cell 3D simulation using artificial neural network for cathodic process description , 2013 .

[47]  J.R. Hockenberry,et al.  Evaluation of uncertainty in dynamic simulations of power system models: The probabilistic collocation method , 2004, IEEE Transactions on Power Systems.

[48]  Youssef M. Marzouk,et al.  Adaptive Smolyak Pseudospectral Approximations , 2012, SIAM J. Sci. Comput..

[49]  Murat Peksen,et al.  A 3D CFD model for predicting the temperature distribution in a full scale APU SOFC short stack under transient operating conditions , 2014 .

[50]  N. Brandon,et al.  Hydrogen production through steam electrolysis: Model-based steady state performance of a cathode-supported intermediate temperature solid oxide electrolysis cell , 2007 .

[51]  Mohsen Assadi,et al.  Artificial neural network model of a short stack solid oxide fuel cell based on experimental data , 2014 .

[52]  Jianzhong Zhu,et al.  Operating limitation and degradation modeling of micro solid oxide fuel cell-combined heat and power system , 2019, Applied Energy.