Online State Of Charge Estimation Research Articles

A Novel Discharge Mode Identification Method for Series-Connected Battery Pack Online State-of-Charge Estimation Over A Wide Life Scale

Lithium-ion batteries are widely used in energy storage nowadays. However, differences caused by aging among in-pack cells are inevitable, which makes accurate state-of-charge (SOC) estimation for packs still challenging. In this article, a novel discharge mode identification (DMI) method for series-connected battery pack online SOC estimation is proposed. The DMI method simplifies the process of searching for “poor SOC cell.” The discharge process is defined into two different modes on the basis of inconsistency analysis. The pack SOC is estimated based on the average cell or all cells during different discharge phases, respectively. Furthermore, considering lower computational load is critical in a battery management system (BMS), a novel segmented coulomb counting (SCC) method based on partial adaptive forgetting factors recursive least square (PAFFRLS) is proposed as a part of the DMI method, which provides a balanced solution to the cell SOC estimation. Eventually, numerous simulations and experiments for LiNCM and LiFePO4 packs are employed to verify the validity of the proposed DMI over a wide life scale. The average estimation errors of a series-connected battery pack under different working conditions at different temperatures are within 2.5%, which shows a good performance and provides a better guidance to the design of BMS.

An online SOC and capacity estimation method for aged lithium-ion battery pack considering cell inconsistency

Abstract For lithium-ion battery packs, especially aged lithium-ion batteries, the inconsistencies in State-of-Charge (SOC), model parameter and capacity between cells cannot be ignored. In order to accurately estimate the SOC and capacity of each cell in the lithium-ion battery pack online, a Special and Difference (S&D) model, i.e. a serial-connected battery pack model, is established based on a second-order equivalent circuit model as cell model. The multi-time scale extended Kalman filter algorithm is proposed based on “S&D” model to estimate the SOC, model parameter and capacity of each cell in the battery pack. The proposed algorithm involves three time dimensions: a short time scale which contains special cell's SOC and model parameter estimation, a middle time scale which contains the remaining cells’ SOC and model parameter estimation, and a long time scale which contains all cells’ capacity estimation. The multi-time scale extended Kalman filter algorithm for aged battery pack is verified under two dynamic conditions. The results show that the SOC estimation error of each cell in the battery pack is within 5% in the whole testing period and it is within 3% when the later capacity estimation process keeps stable. In addition, the number of the cells with maximum and minimum capacity can be accurately identified after the middle stage of the capacity estimation process, which is significant for the consistency management of the battery pack.

Joint estimation of battery state-of-charge and state-of-health based on a simplified pseudo-two-dimensional model

With the advantages of accurate and elaborate, the application of the electrochemical models has become a potential trend for states estimation of Li-ion batteries in advanced battery management systems (BMS). This paper presents an original approach for joint estimation of state-of-charge (SoC) and state-of-health (SoH) based on the pseudo-two-dimensional (P2D) model. First, the standard P2D model is simplified and reformulated into a nonlinear state-space form with guaranteed observability. Next, the particle filter (PF) algorithm is employed to estimate the average lithium concentration in real time for SoC calculation. In addition, the battery SoH is calibrated based on the prediction of the average lithium concentration at the cut-off voltages of charging and discharging, which, in turn, improves the accuracy of online SoC estimation for aged batteries. For validation purposes, fifteen batteries with the aging state from 100% to 70% of initial capacity are tested under dynamic current profiles. Results show that the maximum SoH estimation error can be limited to 2.8%, and the SoC estimation error is bounded by 2% for new and aged batteries with the calibrated SoH.

A normal‐gamma‐based adaptive dual unscented Kalman filter for battery parameters and state‐of‐charge estimation with heavy‐tailed measurement noise

This study simultaneously considers the state-of-charge (SOC) estimation and model parameter identification of lithium-ion batteries with outliers in measurements. Conventional Kalman-type filters may degrade performance in this case since they assume Gaussian-distributed measurement noise. To improve the SOC estimation accuracy under this condition, a robust normal-gamma (NG)-based adaptive dual unscented Kalman filter (NG-ADUKF) is proposed. First, by modeling the joint distribution of the state and auxiliary variables of the measurement noise as the NG distribution, the unscented Kalman filter (UKF) is integrated with the NG filter to deal with the heavy-tailed measurement noise. Second, the online parameter identification and SOC estimation are realized simultaneously by alternatively using two NG-based adaptive UKFs. The performance of the proposed algorithm is validated by the New European Driving Cycle and Urban Dynamometer Driving Schedule tests. Experimental results show that the proposed NG-ADUKF algorithm has more accurate SOC estimations compared with the dual UKF (DUKF) and the variational Bayes-based adaptive DUKF (VB-ADUKF) in the case of mistuning and outliers. Moreover, the proposed method is more computationally efficient than VB-ADUKF.

Comparative Study of the Influence of Open Circuit Voltage Tests on State of Charge Online Estimation for Lithium-Ion Batteries

The accurate state of charge (SoC) online estimation is a significant indicator that relates to driving ranges of electric vehicles (EV). The relationship between open circuit voltage (OCV) and SoC plays an important role in SoC estimation for lithium-ion batteries. To compare with the traditional incremental OCV (IO) test and the low current OCV (LO) test, a novel OCV test which combines IO test with LO test (CIL) is proposed in this paper. Based on the reliable parameters online identification of the dual polarization (DP) battery model, two SoC estimation algorithms are compared on the accuracy, robustness and convergence speed for the entire SoC region. Meanwhile, the comparative study of the three OCV-SoC relationships fits by the corresponding OCV tests is discussed in terms of the SoC online estimation under various temperatures. The results show that the adaptive extended Kalman filter (AEKF) algorithm can better improve the accuracy and robustness of SoC estimation than that of the extended Kalman filter (EKF) algorithm. Most importantly, the OCV-SoC relationship obtains from the CIL OCV test method is applied to the AEKF algorithm, which has higher accuracy and better statistical indices of SoC estimation, especially suitable for the low temperature.

State-of-charge Estimation of Lithium-ion Battery Based Online Parameter Identification

Accurately estimating the state of charge (SOC) of lithium-ion is very important to improving the dynamic performance and energy utilization efficiency. In order to reduce the influence of model parameters and system coloured noise on SOC estimation accuracy, this paper proposes the SOC estimation based on online identification. Based on the mixed simplified electrochemical model, the forgetting factor recursive least squares (FFRLS) method was used to identify the parameters online, and the SOC estimation was carried out in combination with Unscented Kalman Filter (UKF). Finally, the accuracy and feasibility of the method are verified by Federal Urban Driving Schedule (FUDS), the online identification and SOC estimation are carried out. The experimental results show that the SOC estimation of online parameter identification is more accurate, the system stability is faster and the error is smaller.

An online state of charge estimation for Lithium-ion and supercapacitor in hybrid electric drive vehicle

Abstract Hybrid Energy Storage Systems (HESSs), which are mainly based on Lithium-ion ( L i − i o n ) batteries and supercapacitors (SCs), are extensively investigated for large-scale application such as autonomous trucks, autonomous mobile robots, delivery drones and more precisely Hybrid Electric Vehicle (HEV) applications. This hybridization combines the superior energy density characteristics from L i − i o n battery and the quick ability of SC for energy storage with a virtually unlimited number of charge and discharge cycles. The combination of L i − i o n battery with SC adds more HEV autonomy but increases the complexity of the power management system (PMS). Generally, a good estimation for the state of charge (SOC) of these two power sources creates an opportunity for optimizing the PMS and the safety of HEV. This paper presents a real time state of charge estimation for L i − i o n battery (SOCb) and for SC (SOCsc) using an Extended Kalman Filter (EKF). Whereas, a powerful E K F − S O C estimator needs a precise dynamic battery/SC model. So, an hybrid model that benefits from the advantages of direct measurement Open Circuit Voltage (OCV) method and a Recursive Least Square (RLS) with a forgetting factor is used in this work in order to obtain a globally optimal estimating performance. Nevertheless, this work try to simplify the hard task of L i − i o n battery/SC SOC estimation for any modeling needs, just with few preliminary experimental tests. Results show the efficiency of EKF using a R L S − O C V hybrid method for identifying L i − i o n battery/SC parameters. Finally, an implementation of this reliable hybrid method had been established in test HEV platform.

SOC Estimation with an Adaptive Unscented Kalman Filter Based on Model Parameter Optimization

State of charge (SOC) estimation is generally acknowledged to be one of the most important functions of the battery management system (BMS) and is thus widely studied in academia and industry. Based on an accurate SOC estimation, the BMS can optimize energy efficiency and protect the battery from being over-charged or over-discharged. The accurate online estimation of the SOC is studied in this paper. First, it is proved that the second-order resistance capacitance (RC) model is the most suitable equivalent circuit model compared with the Thevenin and multi-order models. The second-order RC equivalent circuit model is established, and the model parameters are identified. Second, the reasonable optimization of model parameters is studied, and a reasonable optimization method is proposed to improve the accuracy of SOC estimation. Finally, the SOC is estimated online based on the adaptive unscented Kalman filter (AUKF) with optimized model parameters, and the results are compared with the results of an estimation based on pre-optimization model parameters. Simulation experiments show that, without affecting the convergence of the initial error of the AUKF, the model after parameter optimization has a higher online SOC estimation accuracy.

A novel approach to reconstruct open circuit voltage for state of charge estimation of lithium ion batteries in electric vehicles

Open circuit voltage (OCV) has a considerable influence on the accuracy of battery state of charge (SOC) estimation. Three efforts have been made to reconstruct OCV for SOC estimation of lithium ion batteries in this study: (1) A new parameter backtracking strategy is proposed for online parameter identification using the recursive least square (RLS) algorithm to obtain stable OCV, which significantly reduces the jitters occurring in OCV identification results. (2) Historical experimental data of lithium ion batteries are used to derive baseline OCV curve and determine constraint boundaries, then an extended Kalman filter (EKF) is employed as a state observer to estimate the SOC for the same types of the batteries that have not been tested. (3) The OCV-SOC curve is reconstructed based on the accumulated online parameter identification and SOC estimation results. The OCV curve can be locally reconstructed even when the accumulated data only cover a partial range of SOC, which is suitable for electric vehicle (EV) operation conditions. Once the OCV curve is reconstructed, the response surface model of OCV-SOC-Capacity is applied to update battery capacity. In this way, the OCV curve can be gradually reconstructed from high SOC to low SOC during battery discharging process. The use of the reconstructed OCV curve to estimate SOC significantly improves the SOC estimation accuracy with the maximum error less than 3% for EV operation conditions.

State of Charge Estimation of a Lithium Ion Battery Based on Adaptive Kalman Filter Method for an Equivalent Circuit Model

Due to its accuracy, simplicity, and other advantages, the Kalman filter method is one of the common algorithms to estimate the state-of-charge (SOC) of batteries. However, this method still has its shortcomings. The Kalman filter method is an algorithm designed for linear systems and requires precise mathematical models. Lithium-ion batteries are not linear systems, so the establishment of the battery equivalent circuit model (ECM) is necessary for SOC estimation. In this paper, an adaptive Kalman filter method and the battery Thevenin equivalent circuit are combined to estimate the SOC of an electric vehicle power battery dynamically. Firstly, the equivalent circuit model is studied, and the battery model suitable for SOC estimation is established. Then, the parameters of the corresponding battery charge and the discharge experimental detection model are designed. Finally, the adaptive Kalman filter method is applied to the model in the unknown interference noise environment and is also adopted to estimate the SOC of the battery online. The simulation results show that the proposed method can correct the SOC estimation error caused by the model error in real time. The estimation accuracy of the proposed method is higher than that of the Kalman filter method. The adaptive Kalman filter method also has a correction effect on the initial value error, which is suitable for online SOC estimation of power batteries. The experiment under the BBDST (Beijing Bus Dynamic Stress Test) working condition fully proves that the proposed SOC estimation algorithm can hold the satisfactory accuracy even in complex situations.

FPGA implementation of extended Kalman filter for SOC estimation of lithium‐ion battery in electric vehicle

AbstractA FPGA implementation for a model‐based state of charge (SOC) estimation is described in this paper. A Thevenin equivalent circuit model is designed for SOC estimation. The extended Kalman filter (EKF) is designed to complete the SOC estimation, and the error is within 1 % . The FPGA is chosen to achieve realtime SOC estimation. A fast matrix method is proposed to improve the calculation speed of the EKF in FPGA because the EKF algorithm requires many matrix operations. In addition, the embedded system based on the FPGA with a system on a programmable chip (SOPC) technique is built using the Qsys platform in Quartus II. Based on the embedded system, an online testing platform is established to monitor the terminal voltage and load current of the experimental battery in real time; experimental results show that the online SOC estimation is successful. The measurement results show that the FPGA embedded scheme of the EKF allows for successful implementation of the SOC estimation with accuracy and speed. The fast matrix method requires 0.00007 s to implement the SOC estimation and is four times faster than the conventional matrix method.

Co-estimation of capacity and state-of-charge for lithium-ion batteries in electric vehicles

Abstract Precise capacity and state-of-charge (SOC) estimations are vital in order to guarantee safe and efficient operation of battery systems in electric vehicles. In this paper, a co-estimation scheme for battery capacity and SOC estimations is proposed, in which an equivalent circuit model (ECM) is used to represent battery dynamics. The recursive least squares (RLS) method and adaptive extended Kalman filter (AEKF) are leveraged simultaneously to achieve online model parameters identification and SOC estimation. Accelerated aging tests are conducted to investigate the relationship between partial voltage curves and aging levels of batteries. The Elman neural network is then employed to realize battery capacity prediction in real-time, which is used to refurbish the actual capacity of battery SOC estimator. The effectiveness of the proposed co-estimation scheme is experimentally verified under different driving cycles at varied temperatures. The results show that the SOC estimation error at room temperature can reach as high as 2% against discrepant aging levels, while the maximum estimation error is within 6% at varied temperatures.

Online state of charge estimation of Li-ion battery based on an improved unscented Kalman filter approach

• State equations are derived from the battery equivalent circuit model. • Improved unscented Kalman filter approach composed of model adaptive and noise adaptive was introduced. • Experiment of sensitivity analysis was designed. • The experimental results revealed the effectiveness of the purposed approach. An improved unscented Kalman filter approach is proposed to enhance online state of charge estimation in terms of both accuracy and robustness. The goal is to address the drawback associated with the unscented Kalman filter in terms of its requirement for an accurate model and a priori noise statistics. Firstly, Li-ion battery modelling and offline parameter identification is performed. Secondly, a sensitivity analysis experiment is designed to verify which model parameter has the greatest influence on state of charge estimation accuracy, in order to provide an appropriate parameter for the model adaptive algorithm. Thirdly, an improved unscented Kalman filter approach, composed of a model adaptive algorithm and a noise adaptive algorithm, is introduced. Finally, the results are discussed, which reveal that the proposed approach’s estimation error is less than 1.79% with acceptable robustness and time complexity.

Power battery parameter online identification for electric vehicle using a decoupling multiple forgetting factors recursive least squares method

Li-ion batteries are widely used in electric vehicles (EVs). However, the accuracy of online SOC estimation is still challenging due to the time-varying parameters in batteries. This paper proposes a decoupling multiple forgetting factors recursive least squares method (DMFFRLS) for EV battery parameter identification. The errors caused by the different parameters are separated and each parameter is tracked independently taking into account the different physical characteristics of the battery parameters. The Thevenin equivalent circuit model (ECM) is employed considering the complexity of battery management system (BMS) on the basis of comparative analysis of several common battery ECMs. In addition, decoupling multiple forgetting factors are used to update the covariance due to different degrees of error of each parameter in the identification process. Numerous experiments are employed to verify the proposed DMFFRLS method. The parameters for commonly used LiFePO 4 (LFP), Li (NiCoMn)O 2 (NCM) battery cells and battery packs are identified based on the proposed DMFFRLS method and three conventional methods. The experimental results show that the error of the DMFFRLS method is less than 15 mV, which is significantly lower than the conventional methods. The proposed DMFFRLS shows good performance for parameter identification on different kind of batteries, and provides a basis for state of charge (SOC) estimation and BMS design of EVs.

Design and implementation of a battery management system with active charge balance based on the SOC and SOH online estimation

the power delivery performance of series and parallel strings connected battery modules or packs is restricted by the worst cells in the string. Each cell has a slightly different capacity and terminal voltage due to manufacturing tolerances and operating conditions. The cell strings tend to loss balance during charging and discharging process. The motivation of this paper is to develop a battery management system (BMS) to monitor and control the temperature, state of charge (SOC) and state of health (SOH) et al. and to increase the efficiency of rechargeable batteries. An active energy balancing system for Lithium-ion battery pack is designed based on the online SOC and SOH estimation. The remainder capacity of the battery is estimated by measuring the terminal voltage for each cell, and the balance system will be triggered when the difference between the SOC of one cell and the average SOC is more or less than a predefined threshold in order to minimize the output voltage ripple. The simulation results indicate that the designed BMS can precisely synchronize the SOC while minimizing the output voltage ripple.

Battery State-of-Charge Estimation Based on Regular/Recurrent Gaussian Process Regression

This paper presents novel machine-learning-based methods for estimating the state of charge (SoC) of lithium-ion batteries, which use the Gaussian process regression (GPR) framework. The measured battery parameters, such as voltage, current, and temperature, are used as inputs for regular GPR, whereas the SoC estimate at the previous sample is fed back and incorporated into the input vector for recurrent GPR. The proposed methods consist of two parts. In the first part, training is performed wherein the optimal hyperparameters of a chosen kernel function are determined to model data properties. In the second part, online SoC estimation is carried out according to the trained model. One of the practical advantages of a GPR framework is to quantify estimation uncertainty and, hence, to enable reliability assessment of the battery SoC estimate. The performance of the proposed methods is evaluated by using a simulated dataset and two experimental datasets, one with constant and the other with dynamic charge and discharge currents. The simulations and experimental results show the superiority of the proposed methods in comparison to state-of-the-art techniques including a support vector machine, a relevance vector machine, and a neural network.

A Neural Network-Based Robust Online SOC and SOH Estimation for Sealed Lead–Acid Batteries in Renewable Systems

Sealed-type lead–acid batteries are most common energy storage devices used with renewable systems. Battery state of charge (SOC) and state of health (SOH) estimation is a crucial part which requires maximum possible accuracy to ensure a secure and long-lasting battery energy storage system by cutting off charging and discharging processes at right time (100–40%). In this research work, a neural network (NN)-based simplified SOC and SOH estimation technique is proposed. Proposed technique provides an accurate online SOC and SOH estimation without battery internal parameter information. This technique only requires real-time battery voltage and current information and very simple mathematical calculations to estimate SOC and SOH, which makes it easy to implement at any low-cost micro-controller unit (MCU). Training data for NN have been acquired by using Arduino mega MCU, voltage sensor circuit, and current sensor. The NN program is designed and trained by backpropagation technique in Arduino mega. The calculated weights are further used for estimation. Terminal voltage $$(V_{\mathrm{t}})$$ and open-circuit voltage $$(V_{\mathrm{oc}})$$ are measured for different charging currents $$(I_{\mathrm{ch}})$$ and discharging currents $$(I_{\mathrm{dch}})$$ . Later on, these data are used for training the NN. Experimental results are provided to prove the preeminence of proposed SOC and SOH estimation technique.

An Overview and Comparison of Online Implementable SOC Estimation Methods for Lithium-Ion Battery

With the popularity of electrical vehicles, the lithium-ion battery industry is developing rapidly. To ensure battery safe usage and to reduce its average lifecycle cost, accurate state of charge (SOC) tracking algorithms for real-time implementation are required for different applications. Many SOC estimation methods have been proposed in the literature. However, only a few of them consider the real-time applicability. This paper classifies the recently proposed online SOC estimation methods into five categories. Their principal features are illustrated, and the main pros and cons are provided. The SOC estimation methods are compared and discussed in terms of accuracy, robustness, and computation burden. Afterward, as the most popular type of model-based SOC estimation algorithms, seven nonlinear filters existing in literature are compared in terms of their accuracy and execution time as a reference for online implementation.

A double-scale and adaptive particle filter-based online parameter and state of charge estimation method for lithium-ion batteries

Obtaining an estimation of the parameters and state of charge (SoC) of a lithium-ion battery is crucial for an electric vehicle. The parameters of a battery model are usually different throughout the battery lifetime. To obtain an accurate SoC and parameters and reduce the computational cost, a double-scale dual adaptive particle filter for online parameters and SoC estimation of lithium-ion batteries is proposed. First, the lithium-ion battery is modeled using the Thevenin model. Second, a double-scale dual particle filter is proposed and applied to the battery parameter and SoC estimation. To improve the accuracy and convergence ability to the initial environmental offset, a double-scale dual adaptive particle filter is proposed. Finally, the effectiveness and applicability of the two algorithms are verified by Lithium Nickel Manganese Cobalt Oxide (NMC) batteries of different ages.

Open circuit voltage and state of charge online estimation for lithium ion batteries

Open circuit voltage (OCV), as a nonlinear function of state of charge (SoC) of lithium ion battery, commonly obtained through offline OCV test at certain ambient temperatures and aging stages. The OCV-SoC relationship may be inaccurate in real application due to the difference in operation conditions. In this paper, the OCV is identified by H infinity filter (HIF) in real operation conditions. Due to the no need to derive the jacobian matrices with unscented Kalman filter (UKF), the identified discrete OCV points are propagated to state estimation process instead of the traditional OCV-SoC nonlinear function. Additionally, the polarization voltage across the polarization capacitor is also passed to state estimation in the form of discrete data points. The experimental results demonstrate that the HIF-UKF can obtain the OCV and SoC in real time.