Data-driven diagnosis of high temperature PEM fuel cells based on the electrochemical impedance spectroscopy: Robustness improvement and evaluation

Utilizing machine learning techniques for data-driven diagnosis of high temperature PEM fuel cells is beneficial and meaningful to the system durability. Nevertheless, ensuring the robustness of diagnosis remains a critical and challenging task in real application. To enhance the robustness of diagnosis and achieve a more thorough evaluation of diagnostic performance, a robust diagnostic procedure based on electrochemical impedance spectroscopy (EIS) and a new method for evaluation of the diagnosis robustness was proposed and investigated in this work. To improve the diagnosis robustness: (1) the degradation mechanism of different faults in the high temperature PEM fuel cell was first analyzed via the distribution of relaxation time of EIS to determine the equivalent circuit model (ECM) with better interpretability, simplicity and accuracy;(2) the feature extraction was implemented on the identified parameters of the ECM and extra attention was paid to distinguishing between the long-term normal degradation and other faults;(3) a Siamese Network was adopted to get features with higher robustness in a new embedding. The diagnosis was conducted using 6 classic classification algorithms—support vector machine (SVM), K-nearest neighbor (KNN), logistic regression (LR), decision tree (DT), random forest (RF), and Naive Bayes employing a dataset comprising a total of 1935 collected EIS. To evaluate the robustness of trained models: (1) different levels of errors were added to the features for performance evaluation;(2) a robustness coefficient (Roubust_C) was defined for a quantified and explicit evaluation of the diagnosis robustness. The diagnostic models employing the proposed feature extraction method can not only achieve the higher performance of around 100% but also higher robustness for diagnosis models. Despite the initial performance being similar, the KNN demonstrated a superior robustness after feature selection and re-embedding by triplet-loss method, which suggests the necessity of robustness evaluation for the machine learning models and the effectiveness of the defined robustness coefficient. This work hopes to give new insights to the robust diagnosis of high temperature PEM fuel cells and more comprehensive performance evaluation of the data-driven method for diagnostic application.

Semi-supervised estimation of capacity degradation for lithium ion batteries with electrochemical impedance spectroscopy

Machine learning-based methods have emerged as a promising solution to accurate battery capacity estimation for battery management systems.However,they are generally developed in a supervised manner which requires a considerable number of input features and corresponding capacities,leading to prohibitive costs and efforts for data collection.In response to this issue,this study proposes a convolutional neural network(CNN)based method to perform end-to-end capacity estimation by taking only raw impedance spectra as input.More importantly,an input reconstruction module is devised to effectively exploit impedance spectra without corresponding capacities in the training process,thereby significantly alleviating the cost of collecting training data.Two large battery degradation datasets encompassing over 4700 impedance spectra are developed to validate the proposed method.The results show that accurate capacity estimation can be achieved when substantial training samples with measured capacities are given.However,the estimation performance of supervised machine learning algorithms sharply deteriorates when fewer samples with measured capacities are available.In this case,the proposed method outperforms supervised benchmarks and can reduce the root mean square error by up to 50.66%.A further validation under different current rates and states of charge confirms the effectiveness of the proposed method.Our method provides a flexible approach to take advantage of unlabelled samples for developing data-driven models and is promising to be generalised to other battery management tasks.

Interpretation of electrochemical impedance spectroscopy(EIS) circuit model for soils

Based on three different kinds of conductive paths in microstructure of soil and theory of electrochemical impedance spectroscopy(EIS), an integrated equivalent circuit model and impedance formula for soils were proposed, which contain 6 meaningful resistance and reactance parameters. Considering the conductive properties of soils and dispersion effects, mathematical equations for impedance under various circuit models were deduced and studied. The mathematical expression presents two semicircles for theoretical EIS Nyquist spectrum, in which the center of one semicircle is degraded to simply the equivalent model. Based on the measured parameters of EIS Nyquist spectrum, meaningful soil parameters can easily be determined. Additionally, EIS was used to investigate the soil properties with different water contents along with the mathematical relationships and mechanism between the physical parameters and water content. Magnitude of the impedance decreases with the increase of testing frequency and water content for Bode graphs. The proposed model would help us to better understand the soil microstructure and properties and offer more reasonable explanations for EIS spectra.

Constituent phases of the passive film formed on 2205 stainless steel by dynamic electrochemical impedance spectroscopy

The passive film formed on 2205 duplex stainless steel（DSS） in 0.5 M NaHCO3＋0.5 M NaCl aqueous solution was characterized by electrochemical measurements,including potentiodynamic anodic polarization and dynamic electrochemical impedance spectroscopy（DEIS）.The results demonstrate that there is a great difference between the passive film evolutions of ferrite and austenite.The impedance values of ferrite are higher than those of austenite.The impedance peaks of ferritic and austenitic phases correspond to the potential of 0.15 and 0.25 V in the low potential range and correspond to 0.8 and 0.75 V in the high potential range.The evolutions of the capacitance of both phases are reverse compared to the evolutions of impedance.The thickness variations obtained from capacitance agree well with those of impedance analysis.The results can be used to explain why pitting corrosion occurs more easily in austenite phase than in ferrite phase.

Study on pitting process of 316L stainless steel by means of staircase potential electrochemical impedance spectroscopy

Pitting corrosion of 316L stainless steel in NaCl solution was investigated by means of staircase potential electrochemical impedance spectroscopy（SPEIS）.The investigation focused on the transition of stainless steel from the passive state to pitting corrosion.Based on the evolution of electrical parameters of the equivalent electrical circuit,it is suggested that the most probable mechanism of pit creation is the film breaking model.The result demonstrates that staircase potential electrochemical impedance spectroscopy is an effective method for the investigation of pitting corrosion.

ELECTROCHEMICAL IMPEDANCE SPECTROSCOPY STUDY OF CORROSION INHIBITION OF MODIFIED LIGNOSULPHONATE FOR CARBON STEEL

The corrosion inhibition for carbon steel in circulating cooling water by modified lignosulphonate has been investigated using electrochemical impedance spectroscopy technique. Results show that the inhibition efficiency of modified lignosulphonate GCL2 is a great improvement on that of lignosulphonate. The maximum inhibition efficiency of GCL2 reaches 99.21% at the concentration of 400mg·L^-1 at 303K. The corrosion inhibition of GCL2 is attributed to forming adsorption film on the metal surface for the electrochemical impedance spectroscopy in GCL2 solution shows more than one time-constant.Moreover,results also indicate that it is more efficient in stirring solution than in still solution for GCL2 because the constant of adsorption in stirring solution is much larger than that in still solution. The adsorption of inhibitor GCL2 follows Langmuir＇s adsorption isotherm.

Electrochemical impedance spectroscopy study on the corrosion of the weld zone of 3Cr steel welded joints in CO_2 environments

The welded joints of 3Cr pipeline steel were fabricated with commercial welding wire using the gas tungsten arc welding （GTAW） technique. Potentiodynamic polarization curves, linear polarization resistance （LPR）, electrochemical impedance spectroscopy （EIS）, scan- ning electron microscopy （SEM）, and energy-dispersive spectrometry （EDS） were used to investigate the corrosion resistance and the growth of a corrosion film on the weld zone （WZ）. The changes in electrochemical characteristics of the film were obtained through fitting of the EIS data. The results showed that the average corrosion rate of the WZ in CO2 environments first increased, then fluctuated, and finally de- creased gradually. The formation of the film on the WZ was divided into three stages： dynamic adsorption, incomplete-coverage layer forma- tion, and integral layer formation.

Fast Evaluation of Degradation Degree of Organic Coatings by Analyzing Electrochemical Impedance Spectroscopy Data

The degradation coefficient is proposed to evaluate the degradation degree of organic coatings by directly anaIyzing the Bode plots of the electrochemical impedance spectroscopy （EIS） data. This paper investigated the degradation of phenolic epoxy coating/tinplate system by EIS and the degradation coefficient value, which correlates well with the results of breakpoint frequency and variation of phase angle at 10 Hz. Furthermore, the degradation process was confirmed by scanning electron microscope （SEM） and scanning probe microscopy （SPM）. It is concluded that degradation coefficient can be used for the fast evaluation of degradation degree of organic coatings in practical appli- cations.

Localized Electrochemical Impedance Spectroscopy Study on the Corrosion Behavior of Fe-Cr Alloy in the Solution with Cl^- and SO_4^(2-)

The corrosion behaviors of Fe-Cr alloy under three different pH values solutions with C1- and SO42- were investigated by localized electrochemical impedance spectroscopy （LEIS） measurements and the corrosion products were analyzed by laser Raman spectrometry. The results show that the high corrosion resistance of Fe-Cr Alloy is attributed to a passive film which is formed more easily when the alloy contains a large quantity of Cr element. However, its corrosion resistance varies in the solutions with different pH values, especially in the initial corrosion. The average impedance values in neutral and alkaline solution are much higher than that in acidic solution because the passive film is more likely to dissolve in the acidic condition. Moreover, the destructive effect of C1- and SO42 ions on the passive film is also demonstrated in corrosion process through the change of the impedance value with the steeping time.

Effect of sulphate-reducing bacteria on the electrochemical impedance spectroscopy characteristics of 1Cr18Ni9Ti

The electrochemical characteristics of 1Cr18Ni9Ti in sulphate-reducing bacteria （SRB） solutions and the biofilm of SRB on the surface of the 1Cr18Ni9Ti electrode were studied by electrochemical, microbiological, and surface analysis methods. Electrochemical impedance spectroscopy （EIS） of 1Cr18Ni9Ti was measured in the solutions with and without SRB at the culture time of 2, 4, 8 d, respectively. The measurement used two test methods, the nonimmersion electrode method and the immersion electrode method. It was found that the polarization resistance （Rp） of 1Cr18Ni9Ti in the solutions without SRB is the greatest for each test method. When using the nonimmersion electrode method, Rp shifts negatively at first and then positively, and the time constant is only one. Although using the immersion electrode method, the Rp shifts positively at first and then negatively, and the time constant also changes when the biofilm forms. The biofilm observed through SEM is with pores. It was demonstrated that SRB has accelerated corrosion action on 1Cr18Ni9Ti. The protection effect of the biofilm on the electrode depends on the compact degree of the film.

Prediction of fuel cell performance degradation using a combined approach of machine learning and impedance spectroscopy

Accurate prediction of performance degradation in complex systems such as solid oxide fuel cells is crucial for expediting technological advancements.However,significant challenges still persist due to limited comprehension of degradation mechanisms and difficulties in acquiring in-situ features.In this study,we propose an effective approach that integrates long short-term memory(LSTM) neural network and dynamic electrochemical impedance spectroscopy(DEIS).This integrated approach enables precise prediction of future evolutions in both current-voltage and EIS features using historical testing data,without prior knowledge of degradation mechanisms.For short-term predictions spanning hundreds of hours,our approach achieves a prediction accuracy exceeding 0.99,showcasing promising prospects for diagnostic applications.Additionally,for long-term predictions spanning thousands of hours,we quantitatively determine the significance of each degradation mechanism,which is crucial for enhancing cell durability.Moreover,our proposed approach demonstrates satisfactory predictive ability in both time and frequency domains,offering the potential to reduce EIS testing time by more than half.

Exfoliation corrosion susceptibility of 8090 Al-Li alloy examined by electrochemical impedance spectroscopy

The exfoliation corrosion susceptibility and electrochemical impedance spectrosc opy(EIS) of rolled and peak-aged 8090 Al-Li alloys in EXCO solution were studi ed,and the EIS after exfoliation was simulated. Once exfoliation occurs,two ca pacitive arcs appear in the EIS at high-mediate frequency and mediate-low freq uency respectively. The exfoliation-attacked alloy surface consists of two part s,an original flat alloy surface and a new inter-face exposed to EXCO solution due to the exfoliation. The capacitance corresponding to the new exfoliation in ter-face increases approximately linearly with time at early exfoliation stage,due to the enlargement of the new inter-face. Then it maintains stable,due to the corrosion product covering on the new inter-face. The exfoliation suscepti bility can be judged through the average slope of the capacitance vs time curve of the early exfoliation stage. This average slope of the rolled 8090 alloy is m uch higher than that of the peak-aged 8090 alloy,accordingly the rolled 8090 a lloy is more susceptible to exfoliation than the peak-aged 8090 alloy.

ELECTROCHEMICAL IMPEDANCE SPECTROSCOPY DURING CORROSION PROCESS OF 8090 Al-Li ALLOY IN EXCO SOLUTION

The corrosion behavior and electrochemical impedance spectroscopy (EIS) features of 8090 Al-Li alloys in EXCO solution were investigated, and the EIS was simulated using an equivalent circuit. At the beginning of immersion in EXCO solution, the EIS is comprised by a depressed capacitive arc at high-mediate frequency and an inductive arc at low frequency, and the inductive component decreases and disappears with im- mersion time. Once exfoliation or severe pitting corrosion is produced, two capacitive arcs appear in the EIS. These two capacitive arcs are originated from the two parts of the corroded alloy surface, the original flat alloy surface and the new inter-face exposed to the aggressive EXCO solution due to the exfoliation or pitting corrosion. Some corrosion development features of 8090 Al-Li alloys in EXCO solution can be obtained through simulated EIS information.

Investigation of exfoliation corrosion of rolled AA8090 Al-Li alloy using electrochemical impedance spectroscopy

The exfoliation morphologies and electrochemical impedance spectroscopy (EIS) features of as received rolled AA8090 Al Li alloy in EXCO solution were studied. The EIS was simulated using an equivalent circuit. The results show that once the exfoliation occurs, the EIS is composed of two capacitive arcs at high frequency and mediate low frequency; among them, the capacitance corresponding to high frequency ( C 1) is originated from original flat alloy surface, while the capacitance corresponding to mediate low frequency ( C 2) from new interface exposed to EXCO solution due to the exfoliation and the ratio of C 2 to C 1 increases with exfoliation degree. It is advanced that the exfoliation degree can be quantitatively judged through this ratio.

基于机器学习分类算法解析EIS数据的有机涂层性能评价方法

目的基于机器学习分类算法快速评估有机涂层的防腐性能。方法通过实验室加速试验模拟涂层真实的退化过程,并根据测得的电化学数据,分析不同退化阶段的等效电路元件参数。随后,采用随机抽样方法获取大量数据,用于机器学习模型训练。通过对比支持向量机(SVM)、k最近邻(k-NN)和随机森林(RF)3种不同的机器学习算法,以及多种输入特征集训练的涂层性能分类器模型的准确率,分析最适合用于涂层性能快速评估的机器学习算法和电化学特征。结果根据不同输入特征训练的k-NN和RF模型均表现出良好的预测效果,而SVM模型的预测效果相对较差。根据不同频率范围训练的分类器模型中,在低频区表现最佳,而在高频区表现较差。结论基于阻抗虚部、虚部+实部和阻抗模值3种输入特征训练的RF分类器模型的预测效果最准确。不同频率区间内,低频区的阻抗特征更能准确表征涂层性能。

How do high-voltage cathode and PEO electrolyte get along well?EIS analysis mechanism&potentiometric control strategy

PEO-based all-solid-state electrolytes are extensively utilized and researched owing to their exceptional safety,low-mass-density,and cost-effectiveness.However,the low oxidation potential of PEO makes the interface problem with the high-voltage cathode extremely severe.In this work,the impedance of PEO-based all-solid-state batteries with high-voltage cathode(NCM811)was studied at different potentials.The Nyquist plots displayed a gyrate arc at low-frequencies for NCM811/PEO interface.Based on the kinetic modeling,it was deduced that there is a decomposition reaction of PEO-matrix in addition to de-embedded reaction of NCM811,and the PEO intermediate product(dehydra-PEO)adsorbed on the electrode surface leading to low-frequency inductive arcs.Furthermore,the distribution of relaxation time shows the dehydra-PEO results in the kinetic tardiness of the charge transfer process in the temporal dimension.Hence,an artificial interface layer(CEI_(x))was modified on the surface of NCM811 to regulate the potential of cathode/electrolyte interface to prevent the high-voltage deterioration of PEO.NCM/CEI_(x)/PEO batteries exhibit capacity retentions of 96.0%,84.6%,and 76.8%after undergoing 100 cycles at cut-off voltages of 4.1,4.2,and 4.3 V,respectively.Therefore,here the failure mechanism of high-voltage PEO electrolyte is investigated by EIS and a proposed solving strategy is presented.

Rate Law for Photoelectrochemical Water Splitting Over CuO

Photocatalytic splitting of water over p-type semiconductors is a promising strategy for production of hydrogen.However,the determination of rate law is rarely reported.To this purpose,copper oxide(CuO)is selected as a model photocathode in this study,and the photogenerated surface charge density,interfacial charge transfer rate constant and their relation to the water reduction rate(in terms of photocurrent)were investigated by a combination of(photo)electrochemical techniques.The results showed that the charge transfer rate constant is exponential-dependent on the surface charge density,and that the photocurrent equals to the product of the charge transfer rate constant and surface charge density.The reaction is first-order in terms of surface charge density.Such an unconventional rate law contrasts with the reports in literature.The charge density-dependent rate constant results from the Fermi level pinning(i.e.,Galvani potential is the main driving force for the reaction)due to accumulation of charge in the surface states and/or Frumkin behavior(i.e.,chemical potential is the main driving force).This study,therefore,may be helpful for further investigation on the mechanism of hydrogen evolution over a CuO photocathode and for designing more efficient CuO-based photocatalysts.

Characteristics of sealed plasma electrolytic oxidation coatings with electrochemical impedance spectroscopy

Plasma electrolytic oxidation （PEO） coatings are prepared on aluminium with graphite powders added into the electrolyte. The scanning electron microscopy （SEM） coupled with an energy dispersive x-ray analysis system （EDX） is used to characterize the surface and the cross-section morphologies of the coatings. The electrochemical impedance spectroscopy （EIS） is used not only to evaluate the corrosion resistance but also to analyse the structure of the coating. Results show that graphite powders are embedded in the PEO coating. The corrosion resistances of both the inner barrier and the outer porous layer are greatly improved, and the EIS could give some valuable detailed information about the coating structure.

Characterization of Formation Kinetics of Self-Assembled Thiol Monolayers on Gold by Electrochemical Impedance Spectroscopy

Self-assembled monolayers of octadecanethiol (ODT) on gold have been studied by electrochemical impedance spectroscopy (EIS). The fractional coverage has been examined as a function of immersion time of Au in ODT deposition solution. The fractional coverage exhibits two distinct adsorption steps: an initial rapid step followed by a slow one. The fractional coverage of ODT monolayer increases sharply from zero to more than 99% of its maximum within the first minute. However. it takes a day for the fractional coverage to approach its final value.

Analysis of 3.4 Ah lithium-sulfur pouch cells by electrochemical impedance spectroscopy

Despite great progress in lithium-sulfur(Li-S) batteries, the electrochemical reactions in the cell are not yet fully understood. Electrode processes, complex interfaces and internal resistance may be characterized by electrochemical impedance spectroscopy(EIS). EIS is a non-destructive technique and easy to apply, though there are challenges in ensuring the reproducibility of measurements and the interpretation of impedance data. Here, we present the impedance behavior of a 3.4 Ah Li-S pouch cell characterized by EIS. The impedance changes were analyzed over the entire depth-of-discharge, depth-of-charge,and at various temperatures. Based on the formation of intermediates during(dis)charging, the changes of resistances are observed. Overall, the increase in temperature causes a decrease in electrolyte viscosity,lowering the surface energy which can improve the penetration of the electrolyte into the electrode pores. Moreover, the effect of superimposed AC current during EIS measurement was analyzed, and the results show the dependence of the charge transfer resistance on superimposed AC current which was lower compared to steady-state conditions and consents with theory.