Charge Curves Research Articles

Method of Predicting SOH and RUL of Lithium-Ion Battery Based on the Combination of LSTM and GPR

The state of health and remaining useful life of lithium-ion batteries are important indicators to ensure the reliable operation of these batteries. However, because they cannot be directly measured and are affected by many factors, they are difficult to predict. This paper presents method of jointly predicting state of health and RUL based on the long short-term memory neural network and Gaussian process regression. This method extracts the batteries’ health factors from the charging curve, selects health factors with more relevance than the setting standard as the characteristic of capacity by the maximum information coefficient method, and establishes the battery aging and remaining useful life prediction models with Gaussian process regression. On this basis, the long short-term memory neural network is used to predict the trend of the change in health factors with the increase in cycles, and the results are input into a Gaussian process regression aging model to predict the state of health. Taking the health factors and state of health as the characteristics of remaining useful battery life, a battery remaining useful life model based on Gaussian process regression is established, and the change trend in the remaining useful life can be obtained by inputting the predicted health factors and state of health. In this study, four battery data sets with different depths of charge were used to verify the accuracy and adaptability of the algorithm. The results show that the proposed algorithm has high accuracy and reliability.

Self-discharge prediction method for lithium-ion batteries based on improved support vector machine

An improved support vector regression (SVR) method is proposed for predicting the self-discharge voltage drop (SDV-drop) in lithium-ion batteries. Multiple features were extracted according to the charge and discharge curves of lithium-ion batteries, and the three features having the strongest correlation with the SDV-drop were identified via grey relational analysis. Then, these three features were assigned different weight parameters to obtain composite features which were input into the improved support vector machine through differential evolution algorithm parameter optimization training. Finally, the improved SVR model was obtained. Model training and testing were performed via a battery charge and discharge experiment and battery static experimental data of a new energy vehicle company, and the results indicated that the proposed method had a higher prediction accuracy than the neural-network model and the Gaussian process regression model.

Deep neural network‐driven in‐situ detection and quantification of lithium plating on anodes in commercial lithium‐ion batteries

AbstractLithium plating seriously threatens the life of lithium‐ion batteries at low temperatures charging conditions, but the onboard detection and quantification of lithium plating are severely hampered by the limited available signals and volatile operating conditions in real scenarios. Herein, we propose a detection method to predict the occurrence and quantification of lithium plating under uncertain conditions by only using constant‐current curves during charging based on deep learning. A deep neural network (DNN) is developed to extract data‐driven features induced by lithium plating from the charge curves, avoiding the challenge of manual feature selection. Only using the most common voltage and current signals as inputs, the network exhibits superior adaptability and accuracy. The detection accuracy of the proposed method is 98.64%, while the quantity of the lithium plating can be accurately predicted with a root‐mean‐square error <4.1712 mg. Moreover, the generalization ability of the proposed method is verified by its reliable detection accuracy under conditions that are not used in the training dataset. The detection accuracy is 92.39% for brand new charging conditions and 95.53% for brand new aging states. This method shortens the detection time that currently takes more than several hours (the widely used differential curve analysis) to milliseconds and eliminates the need for a rigorous testing environment, showing great potential for onboard application in future battery management systems.image

Modeling Fast Charge Protocols to Prevent Lithium Plating in a Lithium-Ion Battery

This paper describes a methodology to quantitatively devise a fast charge protocol to prevent lithium plating in a lithium-ion battery (LIB) cell. A lithium plating line was derived by detecting a voltage plateau that occurred during charging with a high current by differential voltage analysis (DVA). To account for the safe effects at a certain margin from the lithium plating line during fast charge, the novel index was expressed as a margin of safety (MS). Two-dimensional modeling was then performed to predict the charge curves and thermal distributions of LIB cell during various conditions according to the margin of safety. To validate the modeling approach for the design of the fast charge protocol, cycling tests were performed under five fast charge protocols and discharge. The higher the MS of the fast charge protocol, the better the performance and lifetime impact of the LIB cell.

Electrochemical investigation of ion-beam sputter-deposited carbon thin films for Li-ion batteries

The C-rate capability of 230 nm- and 16 nm-thin ion-beam sputter-deposited amorphous carbon films, an interesting class of carbonaceous material for lithium-ion batteries, was investigated up to Li-platting. Stepwise ascending and descending constant Li+ currents after each fifth cycle, followed by hundreds of cycles with the highest current were applied. The carbon films show similar cycling with irreversible losses during the first five cycles, followed by reversible cycling with a capacity close to that of graphite. The capacity is significantly lower at high currents; however, it is restored for subsequent cycling again at low currents. Differential charge and differential capacity curves reveal three Li+ uptake and three Li+ release peaks located between 0 and 3 V. Irreversible as well as reversible Li bonding can be associated with all these peaks. Irreversibly bonded Li can be found at the surface (solid electrolyte interphase) and in the bulk of the carbon films (Li trapping). Reversible Li bonding might be possible inside the carbon films in graphite-like nano-domains and at defects. The thinner film reveals a more pseudo-capacitive cycling behavior, pointing to enhanced Li kinetics.Graphical abstract

Effect of Si-Based Anode Lithiation on Charging Characteristics of All-Solid-State Lithium-Ion Battery

The description of the design, manufacturing technology, and test results of thin-film solid-state lithium-ion batteries with a nanocomposite negative electrode Si@O@Al is given herein. This electrochemical system features the hike on the charging curve plateau, which is interpreted as the change from I–V of the Ti-Si@O@Al contact. The latter is due to the change in the type of silicon conductivity during lithiation, as a result of which the ohmic metal-semiconductor contact proves to be biased in the reverse direction, and the charging current is maintained by minority charge carriers. It is shown that the current-conducting component Si@O@Al is formed by a solid solution a-Si(Al), which has a p-type conductivity. The change in the type of conductivity occurs as a result of silicon compensation through lithiation. It was found that Si@O@Al is nonlinear conductor, which can be considered as a percolation cluster formed by amorphous silicon nanoparticles and molecular clusters of silicon dioxide. The height of the Schottky barrier of the Ti|a-Si(Al) contact and the electron affinity of the a-Si(Al) solid solution were estimated.

The nonlinearities in the galvanostatic charging curves of supercapacitors provide insights into charging mechanisms

Supercapacitors are often charged using constant currents. The capacitance can be determined from the slope of the voltage-time curve if the measured voltage over the supercapacitor increases linearly with time. However, the resulting voltage-time curve is often nonlinear, which may lead one to interpret the capacitance as being either time or voltage dependent. In the current work, systematic experimental studies of the nonlinearity of galvanostatic charging curves as a function of applied current and temperature are undertaken for commercial supercapacitors in the range 1–1000 F. A consistent theory is developed to explain the available data. It is demonstrated that the nonlinearity in the voltage-time curve can be attributed to a constant capacitance in parallel with a resistance, the latter which is inversely proportional to the applied current. The influence of faradaic charge transfer reactions or surface charge reorganization on this parallel resistance is analyzed. The proposed theory is also used to analyze galvanostatic charging data available in the research literature, and the different types of nonlinearities observed provide new insight into the mechanisms occurring during charging of various types of supercapacitors.

Determination of electrochemical impedance of lithium-ion battery from charge curve by wavelet transformation

Lithium-ion batteries (LIBs) are being used as power sources for use in electric vehicles and aircraft, and so on. Early detection technique of battery deterioration is required with the spread adoption of electric vehicles and aircraft. Electrochemical impedance spectroscopy is a powerful method to evaluate the internal state of LIB. We recently have proposed a method to determine the impedance spectra of LIB by wavelet transformation (WT) of the external input and the output signals. In the present paper, the impedance spectra were determined by performing WT on the charge curves of the LIB without an external input signal. A capacitive semicircle and Warburg impedance appeared in the impedance spectrum determined by the WT, which is roughly consistent with the impedance spectrum measured by a frequency response analyzer (FRA). In addition, the new impedance determination method was applied to the evaluation of deteriorated batteries. The diameter of the semicircle increased with fading the battery capacity. Curve fitting was applied to the spectra determined by the WT. The sum of the resistance related to electrode structure, Rc, and the charge transfer resistance, Rct, increased as the capacity loss increased. Moreover, the parameter values were in good agreement with those calculated from the impedance measured using FRA, indicating that this new impedance determination method is useful for evaluation of deteriorated batteries. Furthermore, we emphasize that this method can diagnose the battery deterioration simply by charging the battery.

SOC Estimation Based on Hysteresis Characteristics of Lithium Iron Phosphate Battery

In order to improve the estimation accuracy of the state of charge (SOC) of lithium iron phosphate power batteries for vehicles, this paper studies the prominent hysteresis phenomenon in the relationship between the state of charge and the open circuit voltage (OCV) curve of the lithium iron phosphate battery. Through the hysteresis characteristic test of the battery, the corresponding SOC-OCV data when the battery is charged or discharged from different SOC states are analyzed. According to the approximation trend of the hysteresis main loop curve by the data points, a differential equation model for approximately solving the charge or discharge hysteresis small loop curve under any SOC state is established, and the adjustment parameters of the model are analyzed and debugged in sections. Then, based on the second-order Thevenin equivalent circuit model, the forgetting factor recursive least squares method is used to identify the model parameters online. When deriving the relationship between the OCV and SOC, according to the state of charge and discharge and the current SOC value, the approximate model of the real hysteresis small loop curve in the current state is solved in real time, and the extended Kalman recursion algorithm is substituted to correct the corresponding relationship between the OCV and SOC. Finally, the integrated forgetting factor recursive least squares online parameter identification and extended Kalman filter to correct the SOC-OCV hysteresis relationship in real time considering the hysteresis characteristics are used to complete the real-time estimation of the SOC of the lithium iron phosphate battery. The synthesis algorithm proposed in this paper and the Kalman filter algorithm without considering the hysteresis characteristics are compared and verified under the Dynamic Stress Test (DST) data. Based on the method proposed in this paper, the maximum error of terminal voltage is 0.86%, the average error of terminal voltage is 0.021%, the root mean square error (RMSE) of terminal voltage is 0.042%, the maximum error of SOC estimation is 1.22%, the average error of SOC estimation is 0.41%, the average error of SOC estimation is 0.41%, and the RMSE of SOC estimation is 0.57%. The results show that the comprehensive algorithm proposed in this paper has higher accuracy in both terminal voltage following and SOC estimation.

Parametrization of a Doyle-Fuller-Newman (DFN) Model for a Commercial 37Ah Li-Ion Battery and Validation with Incremental Capacity Analysis

Physico-chemical modeling is considered as a very interesting way of simulating the behavior of a Li-ion battery cell by linking its internal state to its performance and aging. However, an accurate parametrization of such models is necessary for a truly representative outcome. In the case of commercial batteries, the parametrization procedure is hardly ever entirely undergone, leaving room for parameters tweaking and inaccurate simulation results. In the framework of this work, over 70% of the parameters were measured for a Samsung SDI 37Ah Li-ion prismatic battery (NMC111/Graphite) for PHEV (Plug-in Hybrid Electric Vehicle) applications. A Doyle-Fuller-Newman (DFN) Pseudo 2D model is chosen based on the work of Dufour et al.1. The major electrochemical characterization technics and measurement results for every component in the cell will be presented. The cells were opened in the ante-mortem procedure2, materials were identified, geometrical parameters were measured and samples were taken for coin-cell experiments. A small portion of the electrolyte was recovered and studied to conclude that it was made of EC-DMC-EMC at 1:1:2 portions. A similar solution was synthetized in order to conduct ionic conductivity measurements at temperatures ranging from -30 to 60°C. Samples from the separator were taken for both effective ionic conductivity and porosity measurements. The tortuosity of the electrodes was calculated via an EIS (Electrochemical Impedance Spectroscopy) with a blocking-electrodes configuration, and finally the diffusion factor in the active particles was estimated via GITT (Galvanostatic Intermittent Titration Technique) experiments. A very slow current regime was used to lithiate and de-lithiate the active materials in a half-cell experiment for electrode balancing purposes. The model was compiled using COMSOL Multiphysics 5.5, and results were compared to validation experiments conducted on the full-setup prismatic cell. The charge curve was validated under different current and power regimes at 25°C, and the Incremental Capacity (IC) curve showed a reasonable accuracy. With this in hand, the following step would be to couple this performance model to an aging model and to compare the evolution of the IC peaks with the experimental aging.

Low-Error Estimation of Half-Cell Thermodynamic Parameters from Whole-Cell Li-Ion Battery Experiments: Physics-Based Model Formulation, Experimental Demonstration, and an Open Software Tool

Low C-rate charge and discharge experiments, plus complementary differential voltage or differential capacity analysis, are among the most common battery characterization methods. Here, we adapt the multi-species, multi-reaction (MSMR) half-cell thermodynamic model to low C-rate cycling of whole-cell Li-ion batteries. MSMR models for the anode and cathode are coupled through whole-cell charge balances and cell-cycling voltage constraint equations, forming the basis for model-based estimation of MSMR half-cell parameters from whole-cell experimental data. Emergent properties of the whole-cell, like slippage of the anode and cathode lithiation windows, are also computed as cells cycle and degrade. A sequential least-square optimization scheme is used for parameter estimation from low-C cycling data of Samsung 18650 NMC|C cells. Low-error fits of the open-circuit cell voltage (e.g., under 5 mV mean absolute error for charge or discharge curves) and differential voltage curves for fresh and aging cells are achieved. We explore the features (and limitations) of using literature reference values for the MSMR half-cell thermodynamic parameters (reducing our whole-cell formulation to a 1-degree-of-freedom fit), and demonstrate the benefits of opening up the degrees of freedom by letting the MSMR parameters be tailored to the cell under test, within constrained bounds near the half-cell reference values. Bootstrap analysis is performed on each dataset to show the robustness of our fitting to experimental noise and data sampling over the course of 600 cell cycles. The results show which specific MSMR insertion reactions are most responsible for capacity loss in each half-cell and the collective interactions that lead to whole-cell slippage and changes in useable capacity. Open-source software is made available to easily extend this model-based analysis to other labs and battery chemistries.

Optimization of charge curve for the extreme inhibition of growing microstructures during electrodeposition

Optimization of charge curve for the extreme inhibition of growing microstructures during electrodeposition

A fast state-of-health estimation method using single linear feature for lithium-ion batteries

Data-driven methods are commonly used for state of health (SOH) estimation, which is essential to battery energy management. However, complex machine learning models, data gathering, and feature processing hinder its further implementation. A fast SOH estimation method based on linear properties of short-time charging is proposed to overcome these challenges. Only the exceptional single linear health factor (LHF) is required for effective SOH estimation. The LHF is chosen through correlation analysis from short-term feature derived from charging curves. The processing is straightforward. To define the relationship between LHF and SOH, a linear regression model is developed. For the simplicity and effectiveness of the method, it is suitable to be implemented in online applications with low hardware requirements. Finally, experiments show that the SOH estimation method has the highest accuracy of 0.54%, and the biggest estimation error is 2.20%. Furthermore, the data from first 20% cycles of the battery are used to build the model, ensuring that the SOH estimation accuracy is comparable. It is worth noting that the time cost of data acquisition does not exceed 30 s, which is important for fast estimation.

Modified Plett-model for modeling voltage hysteresis in lithium-ion cells

Lithium-ion cells, especially anode materials like graphite or silicon, exhibit a charge-direction dependent voltage hysteresis. In literature, several approaches exist to model this effect. This paper aims at showing modifications to the so called Plett-model. It will be shown that the specific slope of the transition between charge and discharge OCV curves can be linked to the slope of the OCV curves themselves. This leads to a significant reduction in measurement and parameterization time. Furthermore, this paper features the application of the improved model to composite electrodes. The separation of the contributions of graphite and silicon leads to an improved modeling of the total voltage hysteresis.

Fabrication of Eco-friendly Lithium Titanium Phosphate/Lithium Polysulfide/Graphite Rechargeable Batteries

The average global temperature has not changed by more than 1° C over 10,000 years. It is predicted that the global temperature will rise by 3° C by the year 2100, unless action is taken to limit temperature increment to 1.5° C. Since the beginning of the century, the field of lithium-ion battery development has skyrocketed. Li-ion batteries have become the primary electrical energy storage device for portable electronic devices in today's world. This study aims to innovate high-performance, less toxic, low-cost rechargeable lithium-ion batteries with lithium titanium phosphate as anode material, graphite as cathode material and lithium polysulfide as electrolyte under standard laboratory conditions. Cell performance was analyzed by observing the charge and discharge curves. In this research work, the charge-discharge performance was analyzed using LabJack U3-HV with LabJack software. The cell voltage was recorded with time, and using a fixed resistor, current flow was calculated. The cell showed initial voltage of 1.3 V and 10 mA in the first charge curve, the initial voltage was increased to 3.2 V in the second charge curve, and the current has decreased to 5.1 mA. During the discharging process, fast discharging was observed from 2.2 V to 0.2 V, giving current from 3.1 mA to 0.3 mA across 657 Ω. In the charging curve, small voltage plateaus can be seen around 1.5-2.0 V. It was observed that the voltage of charging cycles increases due to the increasement of the cell's internal resistance with time. When increasing the charging voltage across the cell, the current passing through the circuit was decreased by proving Ohm’s law and discharge curves demonstrated fast discharge within 60 minutes.
 Keywords: Li-ion, Lithium titanium phosphate, Charge curve, Discharge curve

Kalman filter and classical Preisach hysteresis model applied to the state of charge battery estimation

The goal of this work is first to include a hysteresis model in the classical equivalent circuit model (ECM) for a battery system and then to improve the estimation of the state of charge (SoC) by applying the Extended Kalman Filter (EKF). The hysteretic behavior of the open circuit voltage (OCV) is modelled with the classical Preisach model used for magnetic materials. The construction of the Preisach operator is made by means of the Everett function identified from experimental data which only involve the charging curves of the battery. Thus, a significant reduction in the time necessary to obtain the measurements is achieved. The model is assessed with some laboratory experiments performed on a lithium-ion battery and the results show that with this procedure hysteresis is very well reproduced, even when interior loops are present. In addition, the use of the EKF allows us to eliminate the measurements noise and ensure the accuracy of SoC estimation. The high computational efficiency and precision of the method, joined to the limited computational resources needed for the numerical implementation, make it particularly suitable for real-time embedded battery management system (BMS) applications. In addition, the proposed methodology is well-adapted to any battery type, independently of the SoC-OCV profile of the hysteresis cycle.

Stress Monitoring of Concrete via Uniaxial Piezoelectric Sensor.

The uniaxial piezoelectric sensor was developed to overcome the problem of reflecting the output charge of the piezoelectric element as a combination of vectors in the three stress directions. The work performance of the uniaxial piezoelectric sensor under varying load patterns and load rates was investigated. The sensor was embedded in concrete to monitor stress, and its elastic modulus was used as the intermediate bridge to establish the correlation between the embedded sensor and the external sensor. Furthermore, a correction factor for the charge transformation strain was suggested to overcome the mismatching of the sensor’s medium and the concrete. Considering related circumstances, a new stress monitoring method based on a uniaxial piezoelectric sensor was proposed, which can achieve stress whole-process monitoring in concrete and confining stress monitoring in the reinforced concrete column. The results reveal that through the proposed method, the output charge curve of the sensor has a substantial overlap with the stress waveform and high fitting linearity. The work performance of the sensor was stable, and its sensitivity was not affected by loading rate and load pattern. The sensor was embedded in concrete and can coordinate with the concrete deformation. The correction factor of strain obtained by the sensor embedded in concrete was 1.07. The relationship between the charge produced by the embedded sensor and its external calibration sensitivity may be used to implement the whole process of stress monitoring in concrete.

Synthesis and electrochemical performance of Mo-doped LiNi0.5Mn1.5O4 cathode material

ABSTRACT The Mo-doped LiNi0.5Mn1.5O4 material has been successfully synthesized by the freezing precipitation method, and the effect of different doping amounts on the properties of LiNi0.5Mn1.5O4 material has been studied. Texting has included XRD, SEM, first charge and discharge curves, cycle performance, rate cycle, CV analysis, and EIS analysis. The experimental results show that the Mo-doping material shows superior performance when the doping amount is 1%, and this amount ensures the highest capacity of the sample. The capacity is 120.1 mAh g−1 at 0.1 C (initial discharge), with which the particle size and the dispersion are uniform. After 50 cycles, the discharge capacity is 118.3 mAh g−1 at 0.1 C, and the capacity retention rate is 98.5%. These results reveal that Mo-doped LiNi0.5Mn1.5O4 contributes to the outstanding electrochemical performance of the LiNi0.5Mn1.5O4.

An Online Estimation Method of State of Health for Lithium-Ion Batteries Based on Constant Current Charging Curve

Accurate estimation of state of health (SOH) is of great significance for the safety and reliability of lithium-ion batteries. In this paper, a novel method to estimate SOH online based on constant current charging curve is presented. In order to incorporate the factor of rates, a simple two-step data transformation process is carried out to make the method suitable for SOH estimation at different charging rates. Then polynomial is used to fit the transformed curve, and the coefficient sets of analytic expression obtained by fitting are taken as the battery aging feature variables. Finally, linear regression algorithm, the simplest machine learning algorithm, is employed to construct the mapping between feature variables and SOH, thus accomplishing the SOH estimation. When estimating SOH, only the charging curve of the whole constant current charging process is needed, regardless of the charging process at whatever rates. This method takes low computational cost, making it suitable for online estimation. The verification results on battery test data show that the method is of high accuracy and effectiveness.

Estimation of maximum available capacity of lithium-ion battery based on multi-view features extracted from reconstructed charging curve

A new method is developed in the paper to estimate the maximum available capacity which is an important basis for indicating the State of Health (SOH) of lithium-ion batteries. Firstly, a data reconstruction approach is proposed to pre-process the acquired data to suppress the influence of measurement noise and reduce the negative impact on estimation precision when measuring equipment adopts different sampling frequencies. Then, the variation trend of the incremental capacity curve obtained based on the reconstructed data with the battery aging is analyzed, and a health indicator (HI) including multi-view features is put forward to characterize the battery degradation more comprehensively. The multi-view features are coming from the capacity increment curve versus voltage and time, including the maximum value of the capacity increment curve, the voltage corresponding to the maximum value, other values surrounding the maximum value and so on. Finally, Support Vector Regression is used to establish a model between the extracted HI and the maximum available capacity, and two types of open source data are used to verify the performance. The experimental results show that the data reconstruction method and multi-view health indicator proposed in the paper can obtain high precision estimation results.