Selection of OCV–SOC Model for Battery Management Systems

Version	Summary	Created by	Modification	Content Size	Created at	Operation
1		Prarthana Pillai	--	1168	2022-09-23 20:17:26	\|
2	format correction	Dean Liu	Meta information modification	1168	2022-09-28 04:22:33	\|

This entry is adapted from the peer-reviewed paper 10.3390/en15186803

A battery management system (BMS) plays a crucial role to ensure the safety, efficiency, and reliability of a rechargeable Li-ion battery pack. State of charge (SOC) estimation is an important operation within a BMS. Estimated SOC is required in several BMS operations, such as remaining power and mileage estimation, battery capacity estimation, charge termination, and cell balancing. The open circuit voltage (OCV) model needs to be stored within the battery management system for real-time SOC estimation. The selection, storage, and processing of OCV models involve several design constraints.

battery management systems Li-ion battery battery model parameter estimation

1. Introduction

Li-ion batteries first entered the commercial market as portable batteries for consumer electronics. Today, the use of battery-operated rechargeable systems is envisioned to be the most promising alternative for hazardous emissions due to the use of fossil fuels ^[1]. Moreover, passenger electric vehicles will continue to see the dominant use of Li-ion batteries ^[2]. In recent times, it has become customary to constantly monitor and manage a battery using a battery management system (BMS) ^[3] to ensures the safe, efficient, and reliable operation of the battery. BMSs are usually made of the following three components: a battery fuel gauge (BFG), an optimal charging algorithm (OCA), and cell balancing circuitry (CBC). The BFG is the most important element of a BMS, and it estimates several important states and parameters of the battery, including the state of charge (SOC). The CBC ensures safety by preventing cell imbalance between batteries in a battery pack. The OCA allows faster charging during usage without affecting the battery’s health. It is important to note that accurate SOC estimation by the BFG is crucial for efficient BMS operation, as both CBC and OCA depend on it. Furthermore, the effect of error in the SOC estimation can also lead to compounded problems such as the reduced lifespan of batteries, over-charging/over-discharging, inefficiency, safety, and reliability issues ^[4]. Thus, research on accurate SOC estimation has intensified over the past decade, and several approaches have been studied for application in BMS.

2. Open-circuit Voltage (OCV) and State of Charge (SOC)

Open-circuit voltage (OCV) is the measure of the electromotive force of the battery. The OCV of a battery is shown to possess a monotonically increasing relationship with the SOC of a battery. Thus, several approaches and models based on the OCV–SOC characterization have been studied for SOC estimation. For this, an OCV–SOC characterization is conducted in a laboratory setting using a scientific-grade battery cycler that is able to maintain precise voltage and current values across the battery terminals. The data collection for the OCV characterization is designed in a way that the effects of the hysteresis and relaxation phenomenons of the battery can be nullified in the obtained OCV model. Depending on the OCV modeling approach, the data collection may also vary. In ^[5], a slow-rate data collection approach is demonstrated on various existing OCV–SOC models for parameter estimation. In this approach, a fully charged battery is very slowly discharged (typically at a C/30 rate) using a constant current until it becomes empty. Then, it is charged back to full charge using the same amount of constant current. This entire discharge–charge process takes 60 h. Constant current ensures that the capacitances of the equivalent circuit model remain saturated; a very low magnitude of current assures that the hysteresis effect can be approximated as an equivalent resistance. By measuring the voltage and current values during this entire discharge/charge process, the OCV–SOC parameters are obtained. It is preferred that these data are free of measurement noise and bias. High-precision battery cyclers can maintain constant currents with very little variation and can measure and store voltage and current with very little measurement noise.

3. Different OCV Models

Different OCV–SOC models exist in the literature to adequately represent the OCV curve in the entire span of SOC (0–100%). Several reasons can be stated as to why many variations in the parametric expression for different models exist. Each model approximates the OCV curve differently for the lower (≈0–30%) and higher (≈80–100%) ranges of SOC. For example, the OCV–SOC relationship is quite approximately a straight line between 30% and 80% of the SOC. The straight-line model is the most simplistic approach to OCV–SOC characterization, needing just two parameters; however, the accuracy of the model is compromised at very low and very high SOC regions. In order to improve the accuracy of representation, higher-order empirical models utilizing special functions, such as the polynomial ^[6], trigonometric ^[7], logarithmic ^[8]^[9]^[10]^[11], and exponential functions ^[12]^[13], are used. The estimated parameters using these special functions often need to be represented up to their nth decimal digit for the precise estimation of SOC (for example, in combined+3 model ^[11], n needs to be as high as six to preserve the modeling accuracy). This directly translates to using a large number of bits to completely represent, store, and process these parameters. However, many practical applications (see examples form Texas Instruments ^[14] and Maxim Integrated ^[15]) only allow low-bit processing for BFGs, requiring the traditional OCV–SOC parameters to be rounded. Rounding has been shown to significantly alter the model representation, resulting in poor SOC estimation accuracy ^[16]. To be able to represent the OCV–SOC curve in low-computing environments precisely, tabular models can be used ^[16]. Finally, variations in battery chemistries are also a driving factor for varied OCV–SOC representations.

4. Selection of OCV–SOC Model

Different OCV–SOC models vary in their formulation, in the methods of estimation of their parameters, and eventually in the resulting SOC estimation error. While accurate SOC estimation is crucial, selecting a model solely based on the accuracy of estimation may not be suitable in many applications. For example, in high-power restrictive medical equipment, such as an implantable cardiac pacemaker, reducing computational complexity is crucial ^[17], while the accuracy of SOC estimation is also important. In the case of electric vehicles (EV), for example, drivers are found to experience range anxiety ^[18]. Here, the accuracy of SOC estimation is crucial, and the computational requirement for SOC estimation is not a concern. These examples illustrate that BMS designers need to take multiple constraints before selecting an OCV–SOC model. There needs to be a systematic approach to selecting a particular OCV–SOC model from the numerous models presented in the literature for a specific application.

The selection of the OCV–SOC model in practical situations is based on requirements that are specific to the application. For example, if high SOC estimation accuracy is required, then models with the lowest error metrics will be selected. This would imply that the computational and memory requirements are high. Most practical situations demand more than one constraint in model selection. The Borda count is an intuitive method for combining different selection criteria for a compromised selection. The Borda count was originally a voting method in which each voter gives a complete ranking of all possible alternatives. Table 1 ranks the OCV models presented based on all the selection criteria discussed.

Table 1. Model selection metrics rankings.

Model	BF	$R^{2}$	ME	RMSE	AIC	AIC2	FPE	BIC	MDL	KL	CosD	C	SR	Rank
(65)	1	1	1	1	1	1	1	1	1	4	4	4	1	1
(14)	3	3	4	3	3	3	3	3	3	3	1	1	14	2
(6)	2	2	2	2	2	2	2	2	2	13	2	2	13	3
(10)	4	4	3	4	4	4	4	4	4	10	3	3	12	4
(7)	5	5	5	5	6	6	6	6	6	5	5	5	6	5
(5)	6	6	7	6	7	7	7	7	7	11	6	6	10	6
(11)	8	8	9	8	8	8	8	8	8	6	7	7	7	7
(8)	7	7	8	7	5	5	5	5	5	14	14	14	11	8
(4)	9	9	11	9	9	9	9	9	9	8	8	8	5	9
(12)	10	10	10	10	10	10	10	10	10	7	9	9	8	10
(13)	11	11	12	11	11	11	11	11	11	9	10	10	9	11
(15)	12	12	13	12	12	12	12	12	12	12	11	11	3	12
(2)	13	13	14	13	13	13	13	13	13	1	13	13	2	13
(3)	14	14	6	14	14	14	14	14	14	2	12	12	4	14

References

Mohammadi, F. Lithium-ion battery State-of-Charge estimation based on an improved Coulomb-Counting algorithm and uncertainty evaluation. J. Energy Storage 2022, 48, 104061.
Krishna, G.; Singh, R.; Gehlot, A.; Akram, S.V.; Priyadarshi, N.; Twala, B. Digital Technology Implementation in Battery-Management Systems for Sustainable Energy Storage: Review, Challenges, and Recommendations. Electronics 2022, 11, 2695.
Espedal, I.B.; Jinasena, A.; Burheim, O.S.; Lamb, J.J. Current trends for state-of-charge (SoC) estimation in lithium-ion battery electric vehicles. Energies 2021, 14, 3284.
Movassagh, K.; Raihan, A.; Balasingam, B.; Pattipati, K. A critical look at coulomb counting approach for state of charge estimation in batteries. Energies 2021, 14, 4074.
Pattipati, B.; Balasingam, B.; Avvari, G.; Pattipati, K.R.; Bar-Shalom, Y. open-circuit voltage characterization of lithium-ion batteries. J. Power Sources 2014, 269, 317–333.
Szumanowski, A.; Chang, Y. Battery Management System Based on Battery Nonlinear Dynamics Modeling. IEEE Trans. Veh. Technol. 2008, 57, 1425–1432.
Relan, R.; Firouz, Y.; Vanbeylen, L.; Timmermans, J.M.; Schoukens, J. Nonparametric analysis of the short-term electrical response of li-ion battery cells. In Proceedings of the 2016 Indian Control Conference (ICC), Hyderabad, India, 4–6 January 2016; pp. 1–6.
Ta, K.P.; Newman, J. Proton Intercalation Hysteresis in Charging and Discharging Nickel Hydroxide Electrodes. J. Electrochem. Soc. 1999, 146, 2769–2779.
Verbrugge, M.; Tate, E. Adaptive State of Charge Algorithm for Nickel Metal Hydride Batteries Including Hysteresis Phenomena. J. Power Sources 2004, 126, 236–249.
Plett, G.L. Extended Kalman filtering for battery management systems of LiPB-based HEV battery packs: Part 2: Modeling and Identification. J. Power Sources 2004, 134, 262–276.
Balasingam, B.; Avvari, G.V.; Pattipati, B.; Pattipati, K.R.; Bar-Shalom, Y. A Robust Approach to Battery Fuel Gauging, Part III: State of Charge Tracking. J. Power Sources 2014. submitted.
Hu, Y.; Yurkovich, S.; Guezennec, Y.; Yurkovich, B.J. Electro-thermal battery model identification for automotive applications. J. Power Sources 2011, 196, 449–457.
Neumann, D.E.; Lichte, S. A Multi-Dimensional Battery Discharge Model with Thermal Feedback Applied to a Lithium-ion Battery Pack. NDIA Ground Vehicle Systems Engineering and Technology Symposium-Modeling & Simulation, Testing and Validation (MSTV) Mini-Symposium. 2011. Available online: http://gvsets.ndia-mich.org/publication.php?documentID=351 (accessed on 15 July 2022).
Texas Instruments. 2022. Available online: https://www.ti.com/power-management/battery-management/fuel-gauges/products.html (accessed on 15 July 2022).
Maxim Integrated. 2022. Available online: https://www.maximintegrated.com/en/products/power/battery-management/MAX17263.html (accessed on 15 July 2022).
Sundaresan, S.; Devabattini, B.; Kumar, P.; Balasingam, B.; Pattipati, K. Tabular open-circuit Voltage modeling of Li-ion Batteries for Robust SOC Estimation. J. Appl. Energy 2022. submitted.
Mostafa, S.; John, E. Reducing power and cycle requirement for fast fourier transform of electrocardiogram signals through low level arithmetic optimizations for cardiac implantable devices. J. Low Power Electron. 2016, 12, 21–29.
Pevec, D.; Babic, J.; Carvalho, A.; Ghiassi-Farrokhfal, Y.; Ketter, W.; Podobnik, V. Electric Vehicle Range Anxiety: An Obstacle for the Personal Transportation (R)evolution? In Proceedings of the 2019 4th International Conference on Smart and Sustainable Technologies (SpliTech), Split, Croatia, 18–21 June 2019; pp. 1–8.

© Text is available under the terms and conditions of the Creative Commons Attribution (CC BY) license; additional terms may apply. By using this site, you agree to the Terms and Conditions and Privacy Policy.

Upload a video for this entry

Information

Subjects: Engineering, Electrical & Electronic

Contributors MDPI registered users' name will be linked to their SciProfiles pages. To register with us, please refer to https://encyclopedia.pub/register :

Prarthana Pillai

Sneha Sundaresan

Pradeep Kumar

Krishna R. Pattipati

Balakumar Balasingam

View Times: 1.4K

Update Date: 28 Sep 2022

Table of Contents

Video Upload Options

Confirm