Using adaptive neuro-fuzzy inference system (ANFIS) for proton exchange membrane fuel cell (PEMFC) performance modeling

2012 ◽  
Vol 26 (11) ◽  
pp. 3701-3709 ◽  
Author(s):  
S. Rezazadeh ◽  
M. Mehrabi ◽  
T. Pashaee ◽  
I. Mirzaee
Keyword(s):  
Fuel Cell ◽  
Fuzzy Inference System ◽  
Proton Exchange Membrane ◽  
Performance Modeling ◽  
Fuzzy Inference ◽  
Proton Exchange ◽  
Inference System ◽  
Neuro Fuzzy ◽  
Exchange Membrane

Application of Adaptive Neuro-Fuzzy Inference System Techniques to Predict Water Activity in Proton Exchange Membrane Fuel Cell

2018 ◽  
Vol 15 (4) ◽  
Author(s):  
Khaled Mammar ◽  
Slimane Laribi
Keyword(s):  
Fuel Cell ◽  
Water Activity ◽  
Fuzzy Inference System ◽  
Proton Exchange Membrane ◽  
Fuzzy Inference ◽  
Proton Exchange ◽  
Operating Conditions ◽  
Inference System ◽  
Neuro Fuzzy ◽  
Exchange Membrane

This work defines and implements a technique to predict water activity in proton exchange membrane fuel cell. This technique is based on the electrochemical impedance spectroscopy (EIS) as sensor and adaptive neuro-fuzzy inference system (ANFIS) as estimator. For this purpose, a proton exchange membrane fuel cell (PEMFC) model has been proposed to study the performances of the fuel cell for different operating conditions where the simulation model for water activity behavior is in the proposed structure. The technique based on ANFIS predicts the PEM fuel cell relative humidity (RH) from the EIS. For creation of ANFIS training and checking database, a new method based on factorial design of experimental is used. To check the proposed technique, the ANFIS estimator will be compared with the output humidity relative observation.

scholarly journals Proton exchange membrane fuel cell degradation prediction based on Adaptive Neuro-Fuzzy Inference Systems

2014 ◽  
Vol 39 (21) ◽  
pp. 11128-11144 ◽  
Author(s):  
R.E. Silva ◽  
R. Gouriveau ◽  
S. Jemeï ◽  
D. Hissel ◽  
L. Boulon ◽  
...  
Keyword(s):  
Fuel Cell ◽  
Proton Exchange Membrane ◽  
Fuzzy Inference ◽  
Proton Exchange ◽  
Fuzzy Inference Systems ◽  
Neuro Fuzzy ◽  
Inference Systems ◽  
Exchange Membrane

Adaptive neuro-fuzzy inference system and artificial neural network modeling of proton exchange membrane fuel cells based on nanocomposite and recast Nafion membranes

2011 ◽  
Vol 37 (4) ◽  
pp. 347-357 ◽  
Author(s):  
Mehdi Amirinejad ◽  
Naser Tavajohi-Hasankiadeh ◽  
Sayed Siavash Madaeni ◽  
Maria Assunta Navarra ◽  
Ezzat Rafiee ◽  
...  
Keyword(s):  
Proton Exchange Membrane ◽  
Fuzzy Inference ◽  
Network Modeling ◽  
Proton Exchange ◽  
Neural Network Modeling ◽  
Inference System ◽  
Artificial Neural Network Modeling ◽  
Neuro Fuzzy ◽  
Nafion Membranes ◽  
Exchange Membrane

scholarly journals Performance Prediction of Proton Exchange Membrane Fuel Cells (PEMFC) Using Adaptive Neuro Inference System (ANFIS)

2020 ◽  
Vol 12 (12) ◽  
pp. 4952 ◽  
Author(s):  
Tabbi Wilberforce ◽  
Abdul Ghani Olabi
Keyword(s):  
Experimental Data ◽  
Proton Exchange Membrane ◽  
Fuzzy Inference ◽  
Proton Exchange ◽  
Ambient Conditions ◽  
Inference System ◽  
Fuel Cell Performance ◽  
Neuro Fuzzy ◽  
Independent Variable ◽  
Exchange Membrane

This investigation explored the performance of PEMFC for varying ambient conditions with the aid of an adaptive neuro-fuzzy inference system. The experimental data obtained from the laboratory were initially trained using both the input and output parameters. The model that was trained was then evaluated using an independent variable. The training and testing of the model were then utilized in the prediction of the cell-characteristic performance. The model exhibited a perfect correlation between the predicted and experimental data, and this stipulates that ANFIS can predict characteristic behavior of fuel cell performance with very high accuracy.

A Nonlinear Fuel Cell Model Based on Adaptive Neuro-Fuzzy Inference System

2013 ◽  
Vol 321-324 ◽  
pp. 1357-1360 ◽  
Author(s):  
Qi Li ◽  
Wei Rong Chen ◽  
Zhi Xiang Liu ◽  
Shu Kui Liu ◽  
Wei Min Tian
Keyword(s):  
Fuel Cell ◽  
Fuzzy Inference System ◽  
Fuzzy Inference ◽  
Learning Algorithm ◽  
Cell Model ◽  
Proton Exchange ◽  
Power Module ◽  
Inference System ◽  
Operational Conditions ◽  
Neuro Fuzzy

A nonlinear model of proton exchange membrane fuel cell (PEMFC) based on an adaptive neuro-fuzzy inference system (ANFIS) is proposed to study different operational conditions effect on the dynamic response of Ballard 1.2kW Nexa power module. A hybrid learning algorithm combining back propagation (BP) and least squares estimate (LSE) is adopted to identify the parameters of input and output membership functions for the improvement of training efficiency in the ANFIS. The comparisons with the experimental data demonstrate that the obtained ANFIS model can efficiently approximate the dynamic output response of Nexa power module and is capable of predicting dynamic performance in terms of stack output voltage with a high accuracy.

Modeling of the Effects of Cyclic Voltammetry (CV) Using Fuzzy Logic with Different Membership Functions for Proton Exchange Membrane Fuel Cell (PEM) with Polyvinyl Alcohol/Nano Silver (PVA/Ag)

2017 ◽  
Vol 16 ◽  
pp. 67-72
Author(s):  
Ali Serhat Ersoyoglu ◽  
Sadik Ata ◽  
Kevser Dincer ◽  
Gürol Önal ◽  
Yusuf Yilmaz
Keyword(s):  
Experimental Data ◽  
Cyclic Voltammetry ◽  
Fuel Cell ◽  
Fuzzy Set ◽  
Proton Exchange Membrane ◽  
Proton Exchange ◽  
Membership Functions ◽  
Inference System ◽  
Significant Difference ◽  
Exchange Membrane

In this study, the effects of cyclic voltammetry (CV) has been modeled with Rule-based mamdani-type fuzzy (RBMTF), by using experimental data for proton exchange membrane fuel cell with PVA/AG. In the system developed, RBMTF apply input parameters are CV, scan rate and time, output parameters are current density and voltage. 12300 values for experimental study also obtained with RBMTF. Membership functions (MFs) are the building blocks of fuzzy set theory, i.e., fuzziness in a fuzzy set is determined by its MF. Accordingly, the shapes of MFs are important for a particular problem since they effect on a fuzzy inference system. They may have different shapes like triangular, trapezoidal, Gaussian, etc. When the results obtained from RBMTF and statistical analyses of experimental data have been compared, it has been determined that the two groups of data are coherent, and that there is not a significant difference between them. As a result, this study indicates that RBMTF with different membership functions can be safely used for CV.

scholarly journals Dynamic Modeling of a Reformed Methanol Fuel Cell System Using Empirical Data and Adaptive Neuro-Fuzzy Inference System Models

2013 ◽  
Vol 11 (2) ◽  
Author(s):  
Kristian K. Justesen ◽  
Søren Juhl Andreasen ◽  
Hamid Reza Shaker
Keyword(s):  
Fuel Cell ◽  
Fuzzy Inference System ◽  
Fuzzy Inference ◽  
Cell Model ◽  
Empirical Methods ◽  
Inference System ◽  
System Models ◽  
Neuro Fuzzy ◽  
The Individual ◽  
Methanol Fuel Cell

In this work, a dynamic matlab Simulink model of an H3-350 reformed methanol fuel cell (RMFC) stand-alone battery charger produced by Serenergy® is developed on the basis of theoretical and empirical methods. The advantage of RMFC systems is that they use liquid methanol as a fuel instead of gaseous hydrogen, which is difficult and energy-consuming to store and transport. The models include thermal equilibrium models of the individual components of the system. Models of the heating and cooling of the gas flows between components are also modeled and adaptive neuro-fuzzy inference system models of the reforming process are implemented. Models of the cooling flow of the blowers for the fuel cell and the burner which supplies process heat for the reformer are made. The two blowers have a common exhaust, which means that the two blowers influence each other's output. The models take this into account using an empirical approach. Fin efficiency models for the cooling effect of the air are also developed using empirical methods. A fuel cell model is also implemented based on a standard model, which is adapted to fit the measured performance of the H3-350 module. All of the individual parts of the model are verified and fine-tuned through a series of experiments and are found to have mean absolute errors between 0.4% and 6.4% but typically below 3%. After a comparison between the performance of the combined model and the experimental setup, the model is deemed to be valid for control design and optimization purposes.

scholarly journals Dynamic Modeling of a Reformed Methanol Fuel Cell System Using Empirical Data and Adaptive Neuro-Fuzzy Inference System Models

2013 ◽  
Author(s):  
Kristian K. Justesen ◽  
Søren J. Andreasen ◽  
Hamid R. Shaker
Keyword(s):  
Fuel Cell ◽  
Fuzzy Inference System ◽  
Fuzzy Inference ◽  
Cell Model ◽  
Empirical Methods ◽  
Inference System ◽  
System Models ◽  
Neuro Fuzzy ◽  
The Individual ◽  
Methanol Fuel Cell

In this work, a dynamic MATLAB Simulink model of a H3-350 Reformed Methanol Fuel Cell (RMFC) stand-alone battery charger produced by Serenergy® is developed on the basis of theoretical and empirical methods. The advantage of RMFC systems is that they use liquid methanol as a fuel instead of gaseous hydrogen, which is difficult and energy consuming to store and transport. The models include thermal equilibrium models of the individual components of the system. Models of the heating and cooling of the gas flows between components are also modeled and Adaptive Neuro-Fuzzy Inference System models of the reforming process are implemented. Models of the cooling flow of the blowers for the fuel cell and the burner which supplies process heat for the reformer are made. The two blowers have a common exhaust, which means that the two blowers influence each other’s output. The models take this into account using an empirical approach. Fin efficiency models for the cooling effect of the air are also developed using empirical methods. A fuel cell model is also implemented based on a standard model which is adapted to fit the measured performance of the H3-350 module. All the individual parts of the model are verified and fine-tuned through a series of experiments and are found to have mean absolute errors between 0.4% and 6.4% but typically below 3%. After a comparison between the performance of the combined model and the experimental setup, the model is deemed to be valid for control design and optimization purposes.

Sign in / Sign up

Export Citation Format

Share Document