Xin-Jian Zhu

Nonlinear modeling of a SOFC stack based on ANFIS identification

Abstract An adaptive neural-fuzzy inference system (ANFIS) model is developed to study different flows effect on the performance of solid oxide fuel cell (SOFC). During the process of modeling, a hybrid learning algorithm combining backpropagation (BP) and least squares estimate (LSE) is adopted to identify linear and nonlinear parameters in the ANFIS. The validity and accuracy of modeling are tested by simulations and the simulation results reveal that the obtained ANFIS model can efficiently approximate the dynamic behavior of the SOFC stack. Thus it is feasible to establish the model of SOFC stack by ANFIS.

Modeling of a proton exchange membrane fuel cell based on the hybrid particle swarm optimization with Levenberg–Marquardt neural network

Abstract This paper presents a nonlinear modeling approach of a proton exchange membrane fuel cell (PEMFC) based on the hybrid particle swarm optimization with Levenberg–Marquardt algorithm neural network (PSO-LM NN). The PSO algorithm converges rapidly during the initial stages of a global search, while it becomes extremely slow around the global optimum. On the contrary, the LM algorithm can achieve faster convergent speed around the global optimum, while it is prone to being trapped in the local minimum. Therefore the hybrid algorithm with a transition from PSO search to LM training is proposed to train the weights and thresholds of neural network, which aims to exploit the advantage of the both algorithms. An accurate mathematical model is an extremely useful tool for the fuel cell design, and neural network is an excellent optional tool for complex nonlinear dynamic system modeling such as PEMFC. In the paper, firstly a highly reduced PEMFC dynamic physical model is established to generate the data for the PSO-LM NN model training and validation, and then the neural network nonlinear autoregressive model based on the PSO-LM algorithm is applied in modeling PEMFC voltage and temperature model, and finally the validation test result demonstrates that the trained PSO-LM NN model can efficiently approach the dynamic behavior of a PEMFC.

Dynamic modeling of SOFC based on a T–S fuzzy model

Abstract The operating temperature and voltage are the key parameters affecting the performance of Solid Oxide Fuel Cell (SOFC). In this article a Takagi–Sugeno (T–S) fuzzy model is proposed to describe the nonlinear temperature and voltage dynamic properties of the SOFC system. During the process of modeling, a Fuzzy Clustering Means (FCM) method is used to determine the nonlinear antecedent parameters, and the linear consequent parameters are identified by a recursive least squares algorithm. The validity and accuracy of modeling are tested by simulations. The simulation results show that it is feasible to establish the dynamic model of SOFC by using the T–S fuzzy identification method.

Modeling of a Fuel Cell Stack by Neural Networks Based on Particle Swarm Optimization

This paper presented a nonlinear voltage modeling procedure of a proton exchange membrane fuel cell (PEMFC) stack by neural networks based on particle swarm optimization (PSO). PEMFC stack is a complex nonlinear system which is hard to model by traditional ways. So neural networks based on particle swarm optimization (PSONN) was developed to identify a nonlinear PEMFC stack voltage model. In the paper, the PSO algorithm trained the connection weights and thresholds of neural networks, and a neural networks nonlinear autoregressive model with exogenous inputs was applied in modeling PEMFC stack voltage model. The simulation indicated that the PSONN model can efficiently approach the behavior of a PEMFC stack.

A Hybrid Experimental Model of a Solid Oxide Fuel Cell Stack

A multivariable hybrid experimental model of a solid oxide fuel cell stack is developed in this paper. The model consists of an improved radial basis function (RBF) neural network model and a pressure-incremental model. The improved RBF model is built to predict the stack voltage with different temperatures and current density. Likewise, the pressure-incremental model is constructed to predict the stack voltage under various hydrogen, oxygen, and water partial pressures. We combine the two models together and make a powerful hybrid multivariable model that can predict the voltage under any current density, temperature, hydrogen, oxygen, and water partial pressure. The validity and accuracy of modeling are tested by simulations, and the simulation results show that it is feasible to build the hybrid multivariable experimental model.

Ce0.8M0.2O2−δ(M=Mn,Fe,Ni,Cu) as SOFC Anodes for Electrochemical Oxidation of Hydrogen and Methane

Oxide anodes such as doped ceria offer improved tolerance for nonidealities in anode environment such as redox cycles, sulfur and other poisons, and hydrocarbons. Mixed-valence transition element in ceria provides an additional redox couple besides Ce 4+ /Ce 3+ in reduced atmosphere, facilitating its electrocatalytic reaction for oxidation of fuels. This paper presents the electrochemical characteristics of Ce 0.8 M 0.2 Ο 2-δ (M =Mn,Fe,Ni,Cu) for oxidation of hydrogen and methane. Ce 0.8 M 0.2 O 2-δ was synthesized, and crystal phase analysis by X-ray diffraction was performed. Single-phase Ce 0.8 M 0.2 O 2-δ (M=Mn,Fe,Ni) were formed. A second phase, CuO, was found in the powders with the nominal composition of Ce 0.8 Cu 0.2 Ο 2-δ Ce 0.8 M 0.2 Ο 2-δ exhibited stability in reducing atmosphere. In comparison, similar microstructural characteristics were found for Ce 0.8 M 0.2 O 2-δ (M = Mn, Fe, Cu). However, Ce 0.8 Ni 0.2 O 2-δ exhibits poor microstructure with large cracks. The electrochemical oxidation of wet hydrogen and wet methane was investigated with impedance spectroscopy by using the three-electrode configuration. It was found that Ce 0.8 M 0.2 O 2-δ (M = Mn, Fe, Ni, Cu) demonstrates relatively low electrochemical activity in both hydrogen and methane. Regarding low n-type conductivity of transition metal cation-containing ceria, it was suggested that an oxide with a high electronic conductivity be added into the Ce 0.8 M 0.2 O 2-δ matrix for improvement of the electrode performance.

Quantitative assessment of retained austenite in quenching and partitioning treated multiphase steel

For multiphase steel, an efficient and accurate quantitative analysis of constituent phases is critical in understanding the mechanical mechanism and optimising its properties. A method based on image digitisation and the concept of categorised linear interception has been proposed to study the amount of retained austenite (RA) residing in different phases in quenching and partitioning treated multiphase steel. It was found that RA at different locations transformed in different strain stages due to considerable influence from the surrounding phases, and the stability of each type of RA was estimated according to the above mentioned method. The proposed multistage transformation was proved to be more accurate than a single stage stability model.

Mechanism of improved hydrogen embrittlement resistance of low-temperature plasma carburised stainless steel

ABSTRACT The present paper focuses on the underlying principles of the effect of S-phase on reducing hydrogen embrittlement susceptibility. Hydrogen permeation test indicates that S-phase formed by low-temperature carburising has minor effect on hydrogen diffusion. Moreover, hydrogen-induced phase transformation caused by hydrogen charging is suppressed in the presence of S-phase. The kernel average misorientation result indicates that lattice distortion introduced by hydrogen charging is suppressed in the presence of S-phase. Cross-sectional EBSD investigations on the surface layer of fractured samples show that S-phase remains extremely stable during tensile deformation after hydrogen charging.

CHAPTER 14:Modelling and Control of Solid Oxide Fuel Cell

Solid oxide fuel cell (SOFC) is a promising energy conversion device with high efficiency, fuel flexibility and reduced emissions. An important tool in fuel cell development is mathematical modeling, which is particularly appropriate for SOFCs, where localized experimental measurements are difficult due to the high operating temperature. This chapter firstly reports static modeling studies of a SOFC using least squares support vector machine (LS-SVM) and genetic algorithm-radial basis function (GA-RBF) neural network, respectively. The development of control systems is an important technology issue in pursuing successful implementation of the SOFC. The results obtained from a good dynamic model can be very useful to guide future research of design, analysis and optimization of the SOFC. Furthermore, this dynamic model can also be used to develop a control system of the SOFC operation. So, an adaptive neural-fuzzy inference system (ANFIS) model and a Hammerstein model are established to describe the nonlinear dynamic properties of the SOFC separately. Finally, to protect the SOFC and meet the voltage demand of DC type loads, a model predictive control (MPC) is developed to control the output voltage of the SOFC. Simulation results demonstrate the potential of the established models and the excellence of the MPC controller.

Model predictive control for MCFC stack operation temperature

Operation temperature of a molten carbonate fuel cell stack should be controlled within a special range in order to improve availability and performance of fuel cells. In this paper, a predictive control algorithm based on the Takagi-Sugeno fuzzy model is developed for the temperature of a molten carbonate fuel cell stack. Based on the future outputs predicted by a Takagi-Sugeno fuzzy model, a discrete optimization of the control action is carried out according to the principle of branch-and-bound method. The simulation results demonstrate that there is a potential to introduce the predictive control based Takagi-Sugeno fuzzy model to the development of fuel cells.