ReviewCritical review of the methods for monitoring of lithium-ion batteries in electric and hybrid vehicles
Introduction
Hybridization and electrification of vehicle propulsion systems have become key trends in the automotive industry in recent years. These trends are considered as primary instruments for increasing the total efficiency and decreasing or even eliminating carbon dioxide (CO2) emissions and other pollutants from vehicles. Batteries form key components not only for pure battery electric vehicles (BEVs) but also for intermediate storage of electrical energy in fuel cell electric vehicles (FCEVs) and other hybrid EVs (HEVs). Currently only lithium-ion batteries (LIBs) are considered as a highly prospective technology for automotive applications because of its good performance characteristics and promising potential for cost reduction [1], [2].
LIB packs are always equipped with a battery management system (BMS). The BMS consists of hardware and software for battery management including, among others, algorithms determining battery states. The continuous determination of battery states during operation is called battery monitoring.
Since batteries are complex electrochemical devices with a distinct nonlinear behavior depending on various internal and external conditions, their monitoring is a challenging task. This task is additionally hindered by considerable changes in battery characteristics over its lifetime due to aging. On the other hand, very precise and especially reliable battery monitoring is a key function of the BMS. This function enables safe and reliable operation of the battery pack and, hence, of the total application where the battery pack is used. Therefore, special algorithms for battery monitoring are required. The requirements on battery monitoring algorithms are summarized in Section 2.
In next sections, the methods for the monitoring of battery states and parameters (Fig. 1) presented in scientific and technical literatures are reviewed.1 The focus is to elaborate their strengths and weaknesses rather than to describe their respective approaches in detail. For detailed descriptions of the methods, the reader is advised to refer to the respective original sources referenced in this work or to the reviews that can be found in Refs. [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22]. In the following sections, only methods that are applied or can be potentially applied for monitoring LIBs in BEVs and HEVs are considered. Some other methods specific for other battery technologies exist, e.g., lead-acid batteries or other applications, but their consideration lies beyond the scope of this paper.
The battery SoC can be employed in the figurative sense as a replacement for a fuel gauge used in conventional vehicles. The SoC is basically the relationship between the residual battery capacity in its present state (Cr) and total capacity Cbat after completely charging the battery, expressed in a percentage: SoC = Cr/Cbat·100%. Methods for the SoC monitoring are considered in Section 3 and methods for the estimation of the total battery capacity in Section 4.
For EV applications, batteries must not only deliver a certain amount of energy to the drive train during operation but also provide a certain power in various situations. The battery's capability to fulfill certain tasks is often referred to as the state of function (SoF). For the energy management system (EMS) operating in EVs, knowing the maximum power that can be applied to and from the battery by charging or discharging, respectively, is essential. This power depends, among others, on the present battery impedance characteristics. The methods for the estimation of the battery impedance and for the prediction of the available power are considered in Sections 5 Methods for the estimation of the battery impedance, 6 Methods for the prediction of the available power, respectively.
The capability of the battery to store energy and provide a certain power decreases over the battery lifetime because of aging. As an indicator for this deterioration, the additional battery state—state of health (SoH)—is defined. The methods for its determination are considered in Section 7.
The final battery state of interest is the remaining useful life (RUL). As RUL usually the remaining time or number of load cycles until the battery reaches its end of life (EoL) is understood. The methods for the RUL estimation are considered in Section 8.
A typical interaction and information flow among individual methods within of the battery monitoring system is shown in Fig. 2.
Section snippets
Requirements on battery monitoring algorithms
In this section the requirements on battery monitoring algorithms are given (Fig. 3). First, the battery monitoring algorithms have to consider battery characteristics. One challenge is that battery characteristics depend significantly on the battery internal and external conditions (for example, SoC, temperature, current). The other challenge is that almost all battery characteristics, including, for example, battery capacity and impedance parameters, changes significantly over the battery
Methods for the SoC estimation
As mentioned in the introduction, the SoC of the battery in EVs can be employed as a replacement for a fuel gauge used in conventional vehicles. Therefore, the determination of the battery SoC is always a part of the BMS. Therefore, there are a wide range of approaches proposed in the literature, most of which are considered in the following sections (see also Fig. 4).
Methods for capacity estimation
The battery capacity is a figure of merit determining the energy that is stored in the battery and is available for usage when the battery is fully charged. The capacity of the particular battery or cell in a new state is defined by the battery or cell design and varies only slightly for individual batteries or cells of a given type because of the production tolerances. Over the battery lifetime, however, the capacity changes considerably due to aging processes. Therefore, the present capacity
Methods for the estimation of the battery impedance
As the battery capacity, the impedance parameters of the battery in a new state are mainly defined by the battery design, but changes significantly over the lifetime due to aging processes. The knowledge of the present battery impedance parameters is mainly required for
- 1)
estimation of the energy losses in the battery during the operation,
- 2)
SoC estimation based on electrical models (Section 3.3.1), and
- 3)
prediction of the available power of the battery.
Especially for the latest task the impedance
Methods for the prediction of the available power
For EV applications, batteries must not only deliver a certain amount of energy to the drive train during operation but also provide a certain power in various situations. The power of the battery can be limited by the voltage, current, SoC, and temperature ranges allowed for safe operation. Since batteries are complex electrochemical devices, their power capability depends on a variety of internal and external conditions: temperature, SoC, and previous load history. In addition, it
Methods for the SoH estimation
The capability of the battery to store energy and provide a certain power decreases over the battery lifetime because of aging. As an indicator for this deterioration, the other battery state—SoH—is defined. The most common understanding is that the battery SoH is 100% when the battery is new and 0% when the capability of the battery to store energy or provide power decreases to a certain minimum level. In HEVs, the SoH related to the battery power capability (SoHP) plays the most important
Methods for the estimation of the remaining useful life (RUL)
Under the parameter of RUL, the remaining time or number of load cycles until the battery reaches a SoH of 0% is understood. There are basically two concepts for the estimation of the battery RUL that can be found in the literature. The first concept (Fig. 8a) is based on the lifetime model used for the SoH estimation, as described in previous section. By using this lifetime model, predicting the battery RUL when the future battery conditions and loads are taken as inputs is possible. To
Conclusion
In this paper, the methods for monitoring of the battery state of charge, capacity, impedance parameters, available power, state of health, and remaining useful life are reviewed with the focus on elaboration of their strengths and weaknesses. To this end, scientific and technical literature is studied and all approaches are classified in various groups.
There are a high number of methods for monitoring of the battery SoC. They differ significantly in the underlying approach, achievable accuracy
References (375)
- et al.
J. Power Sources
(2013) - et al.
J. Power Sources
(2003) - et al.
J. Power Sources
(2005) - et al.
J. Power Sources
(2001) - et al.
- et al.
J. Power Sources
(2011) - et al.
Appl. Energy
(2013) - et al.
J. Power Sources
(2011) - et al.
Appl. Energy
(2013) J. Power Sources
(2004)
J. Power Sources
Appl. Energy
J. Power Sources
Appl. Energy
J. Power Sources
J. Power Sources
J. Power Sources
J. Power Sources
J. Power Sources
Energy Convers. Manage.
J. Power Sources
J. Power Sources
Simul. Model. Pract. Theory
J. Power Sources
Appl. Energy
Technologie-Roadmap Energiespeicher für die Elektromobilität 2030
Produkt-Roadmap Lithium-Ionen-Batterien 2030
Battery Management Systems Design by Modelling
Impedance-Based Battery Monitoring
IEEE Control Syst. Mag.
Moderne Akkumulatoren richtig einsetzen
Advances in Battery Management Using Neural Networks and Fuzzy Logic
Meas. Sci. Technol.
Zustandserkennung von LiFePO4-Batterien für Hybrid- und Elektrofahrzeuge
J. Renew. Sustain. Energy
Adv. Mater. Res.
ISRN Appl. Math.
IEEE Trans. Ind. Electron.
IEEE Trans. Ind. Electron.
Cited by (856)
Single-cell operando SOC and SOH diagnosis in a 24 V lithium iron phosphate battery with a voltage-controlled model
2024, Journal of Energy StorageUnravelling degradation mechanisms and overpotential sources in aged and non-aged batteries: A non-invasive diagnosis
2024, Journal of Energy StorageVoltage effects on lithium extraction/recovery via electrochemical pumping using a La<inf>0.57</inf>Li<inf>0.29</inf>TiO<inf>3</inf> electrolyte
2024, Sustainable Materials and TechnologiesData-driven autoencoder neural network for onboard BMS Lithium-ion battery degradation prediction
2024, Journal of Energy Storage