Detailed explanation of the key technologies of lithium-ion battery management system for electric vehicles

1 Definition of Battery Management System (BMS)

The safe operating area of ​​lithium-ion batteries is shown in Figure 1. The main task of BMS is to ensure the design performance of the battery system: 1) safety; 2) durability; 3) power.

Figure 1 Schematic diagram of safe working area of ​​lithium-ion battery

The basic framework of BMS software and hardware is shown in Figure 2. The functions it should have are: 1) Battery parameter detection. 2) Battery status estimation. 3) Online fault diagnosis. 4) Battery safety control and alarm. 5) Charging control. 6) Battery balancing. 7) Thermal management. 8) Network communication. 9) Information storage. 10) Electromagnetic compatibility.

Figure 2 Basic framework of automotive BMS software and hardware

2 Key technologies of battery management system

2.1 Requirements of battery management system for sensor signals

2.1.1 Single-chip voltage acquisition accuracy

Generally, for safety monitoring, the voltage of each battery string in the battery pack needs to be collected. Different systems have different requirements for accuracy.

Figure 3 Single cell OCV curve and voltage acquisition accuracy requirements

For LMO/LTO batteries, the single cell voltage acquisition accuracy only needs to reach 10 mV. For LiFePO4/C batteries, the single cell voltage acquisition accuracy needs to reach about 1mV. However, the current single cell voltage acquisition accuracy can only reach 5 mV in most cases.

2.1.2 Sampling frequency and synchronization

There are many types of battery system signals, and the battery management system is generally distributed. During the signal acquisition process, there will be synchronization problems with the signals of different control sub-boards, which will affect the real-time monitoring algorithm. When designing a BMS, it is necessary to put forward corresponding requirements for the sampling frequency and synchronization accuracy of the signal.

2.2 Battery State Estimation

The relationship between various battery state estimates is shown in Figure 4. Battery temperature estimation is the basis for other state estimates.

Figure 4 Battery management system algorithm framework

2.2.1 Battery Temperature Estimation and Management

Temperature has a significant impact on battery performance. Currently, only the surface temperature of the battery can be measured, while the internal temperature of the battery needs to be estimated using a thermal model. Thermal management of the battery is performed based on the estimated structure.

Figure 5 Battery internal temperature estimation process

2.2.2 State of Charge (SOC) Estimation

SOC algorithms are mainly divided into single SOC algorithms and fusion algorithms of multiple single SOC algorithms. Single SOC algorithms include ampere-hour integration method, open circuit voltage method, open circuit voltage method based on battery model estimation, and other SOC estimation methods based on battery performance. Fusion algorithms include simple correction, weighting, Kalman filtering, and sliding mode variable structure methods.

SOC estimation methods based on battery models, such as Kalman filtering, are accurate and reliable and are currently the mainstream methods.

2.2.3 State of Health (SOH) Estimation

SOH refers to the degree of deviation between the current performance of the battery and the normal design indicators. Figure 6 is a simple schematic diagram of the battery performance degradation principle. Currently, SOH estimation methods are mainly divided into durability empirical model estimation method and battery model-based parameter identification method.

Figure 6 Lithium-ion battery double water tank model

2.2.4 State of Functionality (SOF) Estimation

Estimating battery SOF can be simply considered as estimating the maximum available power of the battery. Commonly used SOF estimation methods can be divided into two categories: battery MAP-based methods and battery model-based dynamic methods.

2.2.5 Residual Energy (RE) or State of Energy (SOE) Estimation

RE or SOE is the basis for estimating the remaining range of electric vehicles. Compared with the percentage SOE, the application of RE in the actual vehicle range estimation is more intuitive.

Figure 7: Remaining battery energy (RE)

FIG8 is an energy prediction method (EPM) for accurately predicting the remaining discharge energy of a battery suitable for dynamic working conditions.

Figure 8 Structure of battery remaining discharge energy prediction method (EPM)

2.2.6 Fault diagnosis and safe state (SOS) estimation

Fault diagnosis is one of the necessary technologies to ensure battery safety. Safety status estimation is one of the important items of battery fault diagnosis. BMS can give the fault level of the battery according to the safety status of the battery.

2.2.7 Charging Control

Lithium deposition during charging is the main factor affecting battery life. The mechanism of lithium deposition has been studied. Charging management based on lithium deposition status identification will be the main research direction in the future. The charging current should be increased as much as possible and the charging time should be shortened while ensuring that lithium deposition does not occur at the negative electrode of the battery.

2.2.8 Battery consistency and balancing management

The inconsistency of single cells will eventually affect the life of the battery pack, mainly due to the difference in capacity attenuation (irrecoverable) of single cells and the difference in charge. The latter can be compensated by balancing methods.

The battery balancing algorithm is divided into a balancing strategy based on voltage consistency, a balancing strategy based on SOC consistency, and a balancing strategy based on remaining charge capacity. The last balancing algorithm has wider constraints and higher efficiency (Figure 9).

Figure 9 Schematic diagram of dissipative balancing based on remaining charge capacity

3 Conclusion

The basic research methods of lithium-ion battery management system are:

1) Study the mechanism of lithium-ion batteries and gain a deeper understanding of the evolution of battery performance;

2) Test and study the performance of lithium-ion batteries to determine the primary and secondary factors and laws that affect battery performance;

3) Use mechanism-based, semi-empirical or empirical modeling methods to establish a battery system model that can be actually applied in the battery management system;

4) During operation, based on the collectible data, the battery system parameters are identified online or offline to estimate the battery status (SOC, SOH, SOF, SOE and faults), and the vehicle controller is notified through the network to ensure safe and reliable operation of the vehicle.