Search

Existing research on fault diagnosis for polymer electrolyte membrane fuel cells (PEMFC) has advanced significantly, yet performance is hindered by variations in data distributions and the requirement for extensive fault data. In this study, a cross-domain adaptive health diagnosis method for PEMFC is proposed, integrating the digital twin model and transfer convolutional diagnosis model. A physical-based high-fidelity digital twin model is developed to obtain diverse and high-quality datasets for training diagnosis method. To extract long-term time series features from the data, a temporal convolutional network (TCN) is proposed as a pre-trained diagnosis model for the source domain, with feature extraction layers that can be reused to the transfer learning network. It is demonstrated that the proposed pre-trained model can hold the ability to accurately diagnose the various fuel cell faults, including pressure, drying, flow, and flooding faults, with 99.92 % accuracy, through the effective capture of the long-term dependencies in time series data. Finally, a domain adaptive transfer convolutional network (DATCN) is established to improve the diagnosis accuracy across diverse fuel cells by learning domain-invariant features. The results show that the DATCN model, tested on three different target domain devices with adversarial training using only 10 % normal data, can achieve an average accuracy of 98.5 % (30 % improved over traditional diagnosis models). This proposed method provides an effective solution for accurate cross-domain diagnosis of PEMFC devices, significantly reducing the reliance on extensive fault data.

The wind–thermal bundled power system achieves energy complementarity and optimized scheduling, which is an important way to build a new type of energy system. For the safe and stable operation of the wind–thermal bundled power system, accurate data-driven analysis is necessary to maintain real-time balance between electricity supply and demand. By summarizing the development and characteristics of wind–thermal bundled power system in China and different countries, current research in this field can be clearly defined in two aspects: short-term wind power prediction for wind farms and performance evaluation of automatic generation control (AGC) for thermal power generation units. For short-term wind power prediction, it is recommended to focus on historical data preprocessing and artificial intelligence methods. The technical characteristics of different data-driven wind power prediction methods have been compared in detail. For performance evaluation of AGC units, a comprehensive analysis was conducted on current evaluation methods, including the “permitted-band” and “regulation mileage” methods, as well as the issue of evaluation failure in traditional evaluation methods in practical engineering. Finally, the relative optimal dynamic performance of AGC units was discussed and the future trend of data-driven research in wind–thermal bundled power system was summarized.

Cold start is a critical operating scenario for the proton exchange membrane fuel cell (PEMFC), particularly in the field of transportation. Under sub-freezing temperatures, the water inside the cell will freeze and obstruct gas flow paths as well as cover catalyst reaction sites, resulting in a failed startup. This study proposes an optimization method for the -30°C cold start of PEMFC based on a data-driven surrogate model to improve cold start performance and reduce irreversible damage to the cell. A validated PEMFC cold start mechanism model is utilized as the basis for developing an extreme learning machine (ELM) based data-driven surrogate model, which is trained using data collected from the mechanism model and has higher computational efficiency compared with the original model. In addition, the NSGA-II multi-objective optimization algorithm is employed to optimize the current loading strategies and operating parameters using the surrogate model as fitness function. The objectives are to enhance the minimum voltage and reduce startup duration time. Moreover, experimental validation confirms the effectiveness of the proposed method. The test results demonstrate that a cold start from -30°C is achieved within 97 s, with the minimum voltage reaching 0.44 V. Notably, there is a reduction in startup time by 26 s and an increase in the minimum voltage by 0.06 V compared to the base case. This study establishes a foundation for researchers to adjust operating settings during cold start based on diverse applications and requirements.

Transport processes are crucial for the performance of electrochemical energy conversion devices and attract wide attentions. This work focuses on the critical oxygen transport process in the ionomer electrolyte film on a Pt electrode, which highly limits the performance of low Pt-loading fuel cells. Reduction of oxygen transport resistance may be achieved by optimizing the ionomer film morphology. Therefore, the relationship between ionomer film morphology and oxygen transport characteristics is explored by altering ionomer side chain lengths in this work. The results show that the swollen structure with larger water agglomerates in the ionomer film with shorter ionomer side chains is detrimental to the formation of oxygen transport paths. However, the multilamellar structure with an alternating alignment of water agglomerates and PFSA ionomer agglomerates in the ionomer film with longer ionomer side chains has a larger water-PFSA interface, which provides more oxygen transport paths and thus reduces the oxygen transport resistance. This work inspires the novel design concept of the ionomer electrolyte film with low local oxygen transport resistance, i.e., enlarging the water-PFSA interface parallel to the oxygen transport direction via altering the ionomer material properties, which is valuable for the development of low Pt-loading fuel cells.

As a high efficiency hydrogen-to-power device, proton exchange membrane fuel cell (PEMFC) attracts much attention, especially for the automotive applications. Real-time prediction of output voltage and area specific resistance (ASR) via the on-board model is critical to monitor the health state of the automotive PEMFC stack. In this study, we use a transient PEMFC system model for dynamic process simulation of PEMFC to generate the dataset, and a long short-term memory (LSTM) deep learning model is developed to predict the dynamic performance of PEMFC. The results show that the developed LSTM deep learning model has much better performance than other models. A sensitivity analysis on the input features is performed, and three insensitive features are removed, that could slightly improve the prediction accuracy and significantly reduce the data volume. The neural structure, sequence duration, and sampling frequency are optimized. We find that the optimal sequence data duration for predicting ASR is 5 s or 20 s, and that for predicting output voltage is 40 s. The sampling frequency can be reduced from 10 Hz to 0.5 Hz and 0.25 Hz, which slightly affects the prediction accuracy, but obviously reduces the data volume and computation amount.

Energy storage is an effective way to address the instability of renewable energy generation modes, such as wind and solar, which are projected to play an important role in the sustainable and low-carbon society. Economics and carbon emissions are important indicators that should be thoroughly considered for evaluating the feasibility of energy storage technologies (ESTs). In this study, we study two promising routes for large-scale renewable energy storage, electrochemical energy storage (EES) and hydrogen energy storage (HES), via technical analysis of the ESTs. The levelized cost of storage (LCOS), carbon emissions and uncertainty assessments for EESs and HESs over the life cycle are conducted with full consideration of the critical links for these routes. In order to reduce the evaluation error, we use the Monte Carlo method to derive a large number of data for estimating the economy and carbon emission level of ESTs based on the collected data. The results show that lithium ion (Li-ion) batteries show the lowest LCOS and carbon emissions, at 0.314 US$ kWh-1 and 72.76 gCO2e kWh-1, compared with other batteries for EES. Different HES routes, meaning different combinations of hydrogen production, delivery and refueling methods, show substantial differences in economics, and the lowest LCOS and carbon emissions, at 0.227 US$ kWh-1 and 61.63 gCO2e kWh-1, are achieved using HES routes that involve hydrogen production by alkaline electrolyzer (AE), delivery by hydrogen pipeline and corresponding refueling. The findings of this study suggest that HES and EES have comparable levels of economics and carbon emissions that should be both considered for large-scale renewable energy storage to achieve future decarbonization goals.

Proton exchange membrane fuel cell (PEMFC) is regarded as one of the most promising energy conversion devices, but cost and durability are two challenges that hinder its large-scale application. Short-term performance degradation prediction is of great importance for the optimization of operating conditions and maintenance strategies to improve the durability and reduce costs of the PEMFCs. This paper proposes a novel data-driven short-term prediction method combining discrete wavelet transform (DWT), gaussian process regression (GPR) and particle swarm optimization (PSO) algorithm. The DWT decomposes the voltage time series waveform into numerous sub-waveforms with different characteristics. The GPR is utilized to construct individual prediction models for distinct sub-waveforms and the final prediction outcomes are obtained by adding the results of each model. The PSO achieves the automatic hyper-parameters optimization in order to improve the accuracy and robustness of the GPR model. In addition, the effectiveness of the proposed method is fully validated by the degradation datasets of PEMFCs in constant and quasi-dynamic load current conditions as well as real road working conditions. Finally, compared with widely used data-driven models such as artificial neural networks (ANN), support vector machine (SVM), long short-term memory (LSTM) and newly proposed methods in relevant studies, the proposed method exhibits superior accuracy and stability for degradation prediction.

Paper-based microfluidic fuel cells (PMFCs) gradually become alternative energy sources due to their advanced features, including simple structure, low cost, without requiring auxiliary equipment. However, the applications of PMFCs are limited by the poor power density and expensive cost. Therefore, a novel design of catalyst layer with stepped Pd distribution is proposed to improve the PMFC performance and reduce the cost. An experimental investigation is performed to understand the underlying influence mechanism of the catalyst layer properties, including Nafion solution content catalyst layer area, catalyst loading and catalyst distribution. We found that the proportioin choice of 4.76% Nafion solution in preparing catalyst layer achieves a shortest reaction time and a highest output performance. In addition, choosing a small catalyst layer area and high catalyst loading is beneficial to improve the current density. With the hybrid arrangement of stepped catalyst layers, the PMFC reduces the cost and presents highest current density of 44.10 mA cm−2 and peak power density of 12.87 mWcm−2. Finally, our designed PMFC stack is employed to power a thermohygrograph and a blood glucose meter for at least 5 h. These practical applications are in favor of supporting the developments of integrated device with energy generation and electrochemical detection.

Proton exchange membrane fuel cells (PEMFCs) are promising power sources owing to their high-power/energy densities and low pollution emissions. With the increasing demand for electricity for various low-power devices, small-scale storage of electricity encountered bottle-neck, which provides new opportunities for PEMFC. Owing to the high specific energy of PEMFCs, the utilization of this technology for small-scale applications has recently attracted significant attention. In recent years, considerable effort has been made to advance PEMFC technology and applications, especially in the small-scale PEMFC commercial market. The current review provides a holistic overview of the cutting-edge developments of small-scale PEMFCs in the transportation, stationary, and portable power generator fields. In particular, we examine current literature on the development of small-scale PEMFCs and discuss the operating complexity of PEMFC systems for different applications. Finally, we provide a forward-looking perspective on the strategies for small-scale high-specific-power PEMFC systems.

Data-driven modelling methods are being developed in the quest to achieve more accurate performance prediction of protons exchange membrane fuel cell (PEMFC) systems in response to their complicated physicochemical phenomena. However, there is little research in this field detailing the pre-processing and selection of balance of plants (BOP) features for the input layer of system performance prediction at different current densities. Furthermore, most of the previous research applies neural networks based on simulation data rather than real-time bench or vehicle operation datasets which leads to low robustness and unreliable practical results. This paper details the application of a novel algorithm denoted XGBoost-Boruta, which utilises the combination of an ensemble learning approach and a wrapping approach, to improve the robustness of feature selection and to increase the accuracy and robustness of PEMFC system performance prediction. By introduction of the Z score and shadow features to eliminate the randomness of conventional ensemble learning methods, seven key controllable BOP variables of the hydrogen anode, air cathode and cooling subsystems are selected as the original input variables to determine their dependency on the stack voltage. Two case studies are presented for verification and validation of the proposed algorithm based on the real-time dataset of bench experimental data and data obtained from heavy truck operation at current densities ranging from 100 to 1500 mA/cm2. The feature selection strategy, based on the proposed XGBoost-Boruta algorithm, largely decreases the RMSE by 23.8% and 14.1% and the R2 increases by 0.06 and 0.04 of both the bench experimental and the heavy truck validation datasets respectively.

Degradation prediction of proton exchange membrane fuel cell (PEMFC) stack is of great significance for improving the rest useful life. In this study, a PEMFC system including a stack of 300 cells and subsystems has been tested under semi-steady operations for about 931 h. Then, two different models are respectively established based on semi-empirical method and data-driven method to investigate the degradation of stack performance. It is found that the root mean square error (RMSE) of the semi-empirical model in predicting the stack voltage is around 1.0 V, while the predicted voltage has no local dynamic characteristics, which can only reflect the overall degradation trend of stack performance. The RMSE of short-term voltage degradation predicted by the DDM can be less than 1.0 V, and the predicted voltage has accurate local variation characteristics. However, for the long-term prediction, the error will accumulate with the iterations and the deviation of the predicted voltage begins to fluctuate gradually, and the RMSE for the long-term predictions can increase to 1.63 V. Based on the above characteristics of the two models, a hybrid prediction model is further developed. The prediction results of the semi-empirical model are used to modify the input of the data-driven model, which can effectively improve the oscillation of prediction results of the data-driven model during the long-term degradation. It is found that the hybrid model has good error distribution (RSEM = 0.8144 V, R2 = 0.8258) and local performance dynamic characteristics which can be used to predict the process of long-term stack performance degradation.

In this study, a 3-D multiphase model of the proton exchange membrane (PEM) electrolyzer is developed on an open-source computational fluid dynamics (CFD) platform, OpenFOAM. A pseudo-coupling method is presented to consider the detailed two-phase distribution conditions at the interface between the flow channel and liquid/gas diffusion layer (L/GDL) in anode flow channel of a 3-D PEM electrolyzer model. The detailed two-phase distribution conditions can be obtained by high-speed optical images or volume of fluid (VOF) simulation results. Through this method, it is possible to predict the effect of two-phase flow existing in the anode flow channel on the transport phenomena in the porous electrode, and thus cell performance. The 3-D model is validated with experimental test results at three current density conditions (0-1, 0-2, 0-3 A cm−2) considering the existence of two-phase flow. It is found that if the two-phase flow in the anode flow channel is neglected, the simulation polarization curves will deviate from the experimental test results, especially at high current density conditions. In contrast, the simulated polarization curves manage to fit the experimental data accurately with the pseudo-coupling method, indicating that it is necessary to consider the effect of two-phase flow. In addition, this pseudo-coupling model can not only take the multi-phase mass transport into consideration but also the distribution of temperature and current density inside a PEM electrolyzer. With the help of this model, a novel double-layer flow field is proposed to enhance the performance of electrolyzer. It is found that the cell performance of the proposed flow field is 0.171 V better that of the traditional parallel flow field at 3 A cm−2. Also, the temperature distribution and current density distribution of the new flow field are more uniform, which could benefit the durability and performance of PEM electrolyzer.

As a clean electrochemical energy conversion device, proton exchange membrane (PEM) fuel cell will play great effect in the carbon neutral society. It is widely recognized as a challenge to couple the multi-phase flow in gas flow channel and the electrochemical activities in the electrodes in PEM fuel cell modeling. In this study, a two-fluid model is developed to describe the multi-phase flow in the whole cell, in conjunction with the transport phenomena and electrochemical reaction kinetics. We assume the analogy between micro channel and porous media, and derive the gas/liquid velocity ratio for multi-phase flow in channel through Darcy's law for two-phase flow, which is expressed as a function of gas/liquid viscosity ratio and liquid saturation. Integration of the two-fluid model makes it able to predict the two-phase pressure drop and local liquid saturation in gas flow channels in an operating fuel cell. A sub-model of liquid water coverage at the GDL (gas diffusion layer) surface is also developed and incorporated into the model to study the impact of various contact angles at GDL surface on cell performance. The liquid saturation profiles along the flow direction are compared with literature data with reasonable agreement achieved. The model is employed to study a flow field of five parallel channels connected with an inlet/outlet manifold. It shows that multi-phase flow in the gas flow channel leads to more non-uniform flow and oxygen distributions, and hence a larger concentration loss under high current densities than the single-phase channel flow model.

The morphological evolution caused by Ni coarsening is one of the main causes of performance degradation in solid oxide fuel cell (SOFC) anode. In this study, we perform a model-based quantitative analysis to investigate the influences of morphological evolution on SOFC performance and durability. A phase-field model is developed to track the temporal evolution of microstructure as inputs of an electrode numerical model for evaluating the electrochemical performance degradation over time. After model validation, parametric simulations are conducted. It is found that Ni coarsening manifests mainly in two aspects: the reduction of total three-phase boundary (TPB) length, and the lengthening of single TPB. Based on the quantified microstructure parameters, the overpotential increase caused by Ni coarsening is obtained. Besides, the results highlight the importance of Ni content on the temporal evolution. The performance analysis provides an optimal range of Ni content, which can maintain a relatively low overpotential during the whole process.

The development of multi-physics-resolved digital twins of proton exchange membrane fuel cells (PEMFCs) is significant for the advancement of this technology. Here, to solve this scientific issue, a surrogate modelling method that combines a state-of-the-art three-dimensional PEMFC physical model and data-driven model is proposed. The surrogate modelling prediction results demonstrate that the test-set relative root mean square errors (rRMSEs) of the multi-physics fields range from 3.88% to 24.80% and can mirror the multi-physics field distribution characteristics well. In summary, for multi-physics field prediction, the data-driven surrogate model has a comparable accuracy to the comprehensive 3D physical model; however, it considerably reduces the cost of computation and time and achieves the efficient multi-physics-resolved digital-twin. Two model-based designs based on the as-developed digital twin framework, i.e. the PEMFC healthy operation envelope and the PEMFC state map, are demonstrated. This study highlights the potential of combining data-driven approaches and comprehensive physical models to develop the digital twin of complex systems, such as PEMFCs.

Polymer electrolyte membrane (PEM) fuel cells are electrochemical devices that directly convert the chemical energy stored in fuel into electrical energy with a practical conversion efficiency as high as 65%. In the past years, significant progress has been made in PEM fuel cell commercialization. By 2019, there were over 19,000 fuel cell electric vehicles (FCEV) and 340 hydrogen refueling stations (HRF) in the U.S. (~8,000 and 44, respectively), Japan (~3,600 and 112, respectively), South Korea (~5,000 and 34, respectively), Europe (~2,500 and 140, respectively), and China (~110 and 12, respectively). Japan, South Korea, and China plan to build approximately 3,000 HRF stations by 2030. In 2019, Hyundai Nexo and Toyota Mirai accounted for approximately 63% and 32% of the total sales, with a driving range of 380 and 312 miles and a mile per gallon (MPGe) of 65 and 67, respectively. Fundamentals of PEM fuel cells play a crucial role in the technological advancement to improve fuel cell performance/durability and reduce cost. Several key aspects for fuel cell design, operational control, and material development, such as durability, electrocatalyst materials, water and thermal management, dynamic operation, and cold start, are briefly explained in this work. Machine learning and artificial intelligence (AI) have received increasing attention in material/energy development. This review also discusses their applications and potential in the development of fundamental knowledge and correlations, material selection and improvement, cell design and optimization, system control, power management, and monitoring of operation health for PEM fuel cells, along with main physics in PEM fuel cells for physics-informed machine learning. The objective of this review is three fold: (1) to present the most recent status of PEM fuel cell applications in the portable, stationary, and transportation sectors; (2) to describe the important fundamentals for the further advancement of fuel cell technology in terms of design and control optimization, cost reduction, and durability improvement; and (3) to explain machine learning, physics-informed deep learning, and AI methods and describe their significant potentials in PEM fuel cell research and development (R&D).