Tesla electric car has become the industry leader, to a certain extent, thanks to a powerful battery management system. Only to the complex and a large number of batteries for effective control and management, in order to break through the bottleneck of the popularity of electric vehicles. Take a look at how the battery management system.
The main task of BMS is to ensure the performance of the battery system: 1) security; (2) durability; 3) power.
BMS hardware and software should have the function:
1) battery parameter detection.
2) battery state estimation.
3) on line fault diagnosis.
4) battery safety control and alarm.
5) charge control.
6) battery equalization.
7) thermal management.
8) network communication.
9) information storage.
10) electromagnetic compatibility.
2.1 battery management system requirements for sensor signals
2.1.1 monolithic voltage acquisition accuracy
For LMO/LTO battery, the accuracy of single voltage acquisition is only 10 mV. For LiFePO4/C battery, the accuracy of single voltage acquisition is about 1mV. But at present, the voltage acquisition accuracy of single cell can only reach 5 mV.
2.2.1 battery temperature estimation and management
2.1.2 sampling frequency and synchronization
There are a variety of battery system signals, and the battery management system is distributed, the signal acquisition process, different control board signal synchronization problem, will have an impact on the real time monitoring algorithm. When designing the BMS, it is necessary to put forward the corresponding requirements for the sampling frequency and synchronization precision.
2.2 battery state estimation
2.2.2 state of charge (SOC) estimation
SOC algorithm is divided into a single SOC algorithm and a variety of single SOC algorithm fusion algorithm. A single SOC algorithm including the current time integral method, open circuit voltage method, based on open circuit voltage method, the battery model estimates the other battery performance estimation method based on SOC. The fusion algorithm includes simple modification, weighting, Calman filtering and sliding mode variable structure method.
Calman filter based on battery model SOC estimation method is accurate and reliable, is the mainstream method.
2.2.3 health status (SOH) estimation
SOH refers to the current performance of the battery and the deviation of the normal design indicators. Figure 6 is a simple schematic diagram of the battery performance degradation. At present, the SOH estimation methods are mainly divided into two parts: durability empirical model estimation and parameter identification based on battery model.
2.2.4 functional state (SOF) estimation
The estimated battery SOF can be simply considered to be the maximum available power in estimating the battery. The commonly used SOF estimation methods can be divided into two categories: the method based on battery MAP diagram and the dynamic method based on battery model.
2.2.5 residual energy (RE) or energy state (SOE) estimation
RE or SOE is the basis of the remaining mileage estimation of electric vehicles, compared with the percentage of SOE, the application of RE in the actual vehicle mileage estimation is more intuitive.
2.2.6 fault diagnosis and safe state (SOS) estimation
Fault diagnosis is one of the necessary technologies to ensure the safety of the battery. Security state estimation is one of the most important items in battery fault diagnosis. BMS can give the battery fault level according to the safety state of the battery.
2.2.7 charge control
Analysis of charging lithium is the main reason affecting the battery life, the mechanism analysis of lithium has been studied, based on the analysis of lithium charging management state identification is a major research direction in the future, should be to ensure no lithium battery anode analysis case, as far as possible increase the charging current, shorten the charging time.
2.2.8 battery consistency and balance management
The inconsistency of the single cell will ultimately affect the life of the battery, which is mainly caused by the difference in the capacity of the battery capacity (non recovery) and the difference between the two charges. The latter can be compensated by a balanced approach.
The battery equalization algorithm is based on the voltage based equalization strategy, the SOC based equalization strategy and the equilibrium strategy based on the remaining charge. The last algorithm has a wide range of constraints and high efficiency.
The basic research methods of Li ion battery management system:
4) in the process of operation, according to the data collected, using online or offline identification system to estimate the parameters of the battery, the battery status (SOC, SOH, SOF, SOE, and fault) and through the network to inform vehicle controller to ensure the safety of vehicles and reliable operation.