4.7 Article

Performance prediction of proton-exchange membrane fuel cell based on convolutional neural network and random forest feature selection

Related references

Note: Only part of the references are listed.
Article Electrochemistry

A quasi 2D model for the interpretation of impedance and polarization of a planar solid oxide fuel cell with interconnects

Alessandro Donazzi et al.

Summary: A physical model of a high temperature solid oxide fuel cell (HT-SOFC) is presented, predicting polarization curves and impedance spectra while studying temperature, concentration, and current density gradients within the cell structure. Local impedance spectra reveal unique features and question the feasibility of spatially resolved impedance measurements at a fine scale, highlighting the importance of interphase mass transport and the impact of external diffusion on polarization performance.

ELECTROCHIMICA ACTA (2021)

Article Chemistry, Physical

Proton exchange membrane fuel cell performance prediction using artificial neural network

Tabbi Wilberforce et al.

Summary: The study investigates the use of neural networks to predict voltage and current of proton exchange membrane fuel cells, suggesting that the performance of group method data handling neural network is superior to feed forward back propagation neural network. It is proposed that increasing reactant flow rate directly affects cell performance, but should be lower compared to oxygen flow rate for reduced irreversibilities and operating costs.

INTERNATIONAL JOURNAL OF HYDROGEN ENERGY (2021)

Article Energy & Fuels

Deep learning based prognostic framework towards proton exchange membrane fuel cell for automotive application

Jian Zuo et al.

Summary: The combination of PHM techniques with deep neural network approaches holds promise for improving the durability of fuel cell devices and extracting useful features more efficiently. The proposed attention-based RNN model demonstrates high prediction accuracy in forecasting the output voltage degradation of PEMFC, showing potential for enhancing prediction accuracy and implementing PHM in the fuel cell system.

APPLIED ENERGY (2021)

Article Energy & Fuels

Performance prediction of fuel cells using long short-term memory recurrent neural network

Lu Zheng et al.

Summary: The article proposes a performance prediction method for PEMFC using LSTM recurrent neural network, showcasing its accuracy through experimental validation with polarization and voltage degradation curves. The LSTM method outperforms traditional RNN and BPNN in predicting PEMFC performance, with high accuracy and close correlation to experimental data.

INTERNATIONAL JOURNAL OF ENERGY RESEARCH (2021)

Article Chemistry, Physical

3D non-isothermal dynamic simulation of high temperature proton exchange membrane fuel cell in the start-up process

Jun Zhang et al.

Summary: High temperature proton exchange membrane fuel cell (HT-PEMFC) with phosphoric acid doped polybenzimidazole (PBI) electrolyte offers advantages over conventional PEMFC, but starting it from room temperature to optimal operating range is a challenge. A startup strategy and a non-isothermal dynamic model were proposed and developed to investigate startup time and temperature distribution.

INTERNATIONAL JOURNAL OF HYDROGEN ENERGY (2021)

Article Energy & Fuels

Prediction of electrochemical characteristics of practical-size solid oxide fuel cells based on database of unit cell performance

Masashi Kishimoto et al.

Summary: A novel numerical methodology is proposed for predicting the electrochemical characteristics of a practical-size solid oxide fuel cell (SOFC). The method involves dividing the cell into independent one-dimensional unit cells and evaluating the cell's performance based on the performance of local unit cells. By constructing a database of unit cell performance and using theoretical expressions, the method accurately predicts the macroscopic characteristics of the entire cell.

APPLIED ENERGY (2021)

Article Chemistry, Physical

A multi-objective parametric study of the claw hydrogen pump for fuel cell vehicles using taguchi method and ANN

Pengtai Gu et al.

Summary: The optimization of a hydrogen circulation pump is crucial for the performance improvement of Fuel Cell Vehicles (FCVs), with factors like pressure ratio, axial clearance, and rotating speed having significant impacts on volumetric efficiency and shaft power. The study reveals the key factors influencing the performance of a claw-type hydrogen pump.

INTERNATIONAL JOURNAL OF HYDROGEN ENERGY (2021)

Article Chemistry, Multidisciplinary

Meso-macroporous Carbons Decorated with Ample Nitrogen Sites as Bifunctional Catalysts in CO2 Catalytic Conversion and Oxygen Reduction Reaction

Jing Liu et al.

Summary: The study presents an anti-CO2 poisoning catalyst for oxygen reduction reaction (ORR) derived from fructose and g-C3N4, with NaCl as a pore-forming template and structure protective agent. The resulting 3D-NHPC-x catalyst exhibits high electrocatalytic activity, long-term stability, and methanol tolerance in alkaline electrolyte, providing a potential route for producing high-performance and anti-CO2 poisoning ORR catalysts. Additionally, the 3D-NHPC-x catalyst can be used for efficient CO2 capture and catalytic conversion, outperforming various materials in the field.

CHEMISTRYSELECT (2021)

Article Chemistry, Physical

Degradation prediction model for proton exchange membrane fuel cells based on long short-term memory neural network and Savitzky-Golay filter

Bin Zuo et al.

Summary: This paper proposes a PEMFC degradation prediction model based on LSTM and Savitzky-Golay filter. By selecting monitoring parameters, smoothing data, generating training samples with sliding time window, and adopting dropout layer and mini-batch method to improve model generalization, the model accurately predicts fuel cell degradation.

INTERNATIONAL JOURNAL OF HYDROGEN ENERGY (2021)

Article Computer Science, Artificial Intelligence

Predicting energy cost of public buildings by artificial neural networks, CART, and random forest

Marijana Zekic-Susac et al.

Summary: The paper explores modeling the cost of energy consumed in public buildings using three machine learning methods, finding that the hybrid machine learning approach integrating random forest and Boruta method outperforms individual methods in prediction accuracy, and extracts important predictors of energy cost in public buildings.

NEUROCOMPUTING (2021)

Article Chemistry, Multidisciplinary

Designing AI-Aided Analysis and Prediction Models for Nonprecious Metal Electrocatalyst-Based Proton-Exchange Membrane Fuel Cells

Rui Ding et al.

ANGEWANDTE CHEMIE-INTERNATIONAL EDITION (2020)

Article Computer Science, Information Systems

Building Auto-Encoder Intrusion Detection System based on random forest feature selection

XuKui Li et al.

COMPUTERS & SECURITY (2020)

Article Engineering, Electrical & Electronic

A deep reinforcement learning method for managing wind farm uncertainties through energy storage system control and external reserve purchasing

J. J. Yang et al.

INTERNATIONAL JOURNAL OF ELECTRICAL POWER & ENERGY SYSTEMS (2020)

Article Green & Sustainable Science & Technology

Feature extraction and genetic algorithm enhanced adaptive deep neural network for energy consumption prediction in buildings

X. J. Luo et al.

RENEWABLE & SUSTAINABLE ENERGY REVIEWS (2020)

Review Computer Science, Artificial Intelligence

A review of Convolutional-Neural-Network-based action recognition

Guangle Yao et al.

PATTERN RECOGNITION LETTERS (2019)

Article Computer Science, Artificial Intelligence

Batch-normalized deep neural networks for achieving fast intelligent fault diagnosis of machines

Jinrui Wang et al.

NEUROCOMPUTING (2019)

Article Chemistry, Multidisciplinary

Nanoscale Structure Design for High-Performance Pt-Based ORR Catalysts

Meiling Liu et al.

ADVANCED MATERIALS (2019)

Review Thermodynamics

A comprehensive review on hybrid power system for PEMFC-HEV: Issues and strategies

Xueqin Lu et al.

ENERGY CONVERSION AND MANAGEMENT (2018)

Article Multidisciplinary Sciences

Unveiling the high-activity origin of single-atom iron catalysts for oxygen reduction reaction

Liu Yang et al.

PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA (2018)

Article Computer Science, Artificial Intelligence

Automated epileptic seizure detection using improved correlation-based feature selection with random forest classifier

Md Mursalin et al.

NEUROCOMPUTING (2017)

Article Multidisciplinary Sciences

Iron-based cathode catalyst with enhanced power density in polymer electrolyte membrane fuel cells

Eric Proietti et al.

NATURE COMMUNICATIONS (2011)