Battery, what you must know about SOC
As we all know, the core part of electric vehicles is the power battery, and the importance of the power battery is self-evident. The SOC display of the power battery is a key part of the power battery management work.
- Definition of SOC
SOC (State of charge), which is the state of charge, is used to reflect the remaining capacity of the battery. Its value is defined as the ratio of the remaining capacity to the battery capacity, commonly expressed as a percentage. Its value range is 0~1. When SOC=0, it means the battery is completely discharged. When SOC=1, it means the battery is fully charged.
Battery SOC cannot be measured directly, and its size can only be estimated through parameters such as battery terminal voltage, charge and discharge current, and internal resistance. These parameters are also affected by various uncertain factors such as battery aging, ambient temperature changes, and vehicle driving conditions. Therefore, accurate SOC estimation has become an urgent problem to be solved in the development of electric vehicles.
- Overview of SOC estimation methods
Accurate estimation of battery SOC, on the one hand, comes from the requirements of electric vehicles, and efficient management of batteries from the two perspectives of giving full play to battery capabilities and improving safety; on the other hand, the highly nonlinear behavior of electric vehicle batteries during use makes Accurately estimating SOC is difficult. The combination of the two aspects makes the selection of electric vehicle battery SOC estimation method particularly important. Electric experts have sorted out many methods for estimating SOC, including the traditional discharge test method, ampere-hour measurement method, battery internal resistance method, open circuit voltage method, load voltage method, as well as the more innovative Kalman filter method, Various estimation methods, such as fuzzy logic theory method and neural network method, have their own advantages and disadvantages.
3.Analysis of main SOC estimation methods
(1) Discharge test method
The discharge test method is to continuously discharge the target battery with constant current until the battery’s cut-off voltage. The time taken for this discharge process is multiplied by the magnitude of the discharge current, which is regarded as the remaining capacity of the battery. This method is generally used as a calibration method for battery SOC estimation or for post-battery maintenance work. This method is used when the battery SOC value is not known. It is relatively simple, reliable, and the results are relatively accurate. At the same time, it is suitable for different types of batteries. All valid. However, the discharge test method also has two shortcomings: first, the test process of this method takes a lot of time; second, when using this method, the target battery needs to be removed from the electric vehicle, so this method cannot be used to calculate the Power battery in working condition.
(2) Ampere-hour measurement method
Ampere hour measurement method (ampere hour, referred to as AH), also known as current integration method, ampere hour integration method, the principle of ampere hour measurement method is to equate the discharge capacity of the battery under different currents to the discharge capacity under a specific current , the main idea is the Peukert equation. From this, the following equivalent discharge capacity formula is obtained:
The ampere-hour metering method is a relatively simple method to calculate battery SOC. This method only focuses on the external characteristics of the system. During the power estimation process, it only cares about the power flowing into and out of the battery. The ampere-hour measurement method uses the integration method to calculate the charging and discharging capacity of the battery in real time. By recording and calculating the battery’s power for a long time, the remaining power of the battery at a certain moment can finally be obtained. This method is easy to implement, but because the relationship between battery SOC and charge and discharge capacity is not obtained from the inside of the battery, it only records charge and discharge capacity, which will lead to battery SOC cumulative error and low accuracy of the results. Moreover, this method cannot determine the initial value of the battery. Comprehensive consideration of the influencing factors of battery SOC and power compensation can appropriately improve the accuracy of the ampere-hour measurement method.
(3) Open circuit voltage method
The open circuit voltage method is based on the changing relationship between the battery’s open circuit voltage (OpenCircuitVoltage, OCV) and the lithium ion concentration inside the battery, and indirectly fits the one-to-one correspondence between it and the battery SOC. In actual operation, the battery needs to be fully charged and discharged at a fixed discharge rate (usually 1C) until the battery stops discharging at the cut-off voltage. Based on this discharge process, the relationship curve between OCV and SOC is obtained. When the battery is in actual working condition, the current battery SOC can be obtained by looking up the OCV-SOC relationship table based on the voltage value at both ends of the battery. Although this method is effective for all kinds of batteries, it also has its own shortcomings: first, the target battery must be left standing for more than 1 hour before measuring OCV, so that the electrolyte inside the battery can be evenly distributed to obtain a stable terminal voltage; second, the battery is at different temperatures Or in different life periods, although the open circuit voltage is the same, the actual SOC may be quite different. The measurement results of long-term use of this method cannot guarantee complete accuracy. Therefore, the open circuit voltage method, like the discharge test method, is not suitable for estimating battery SOC during operation.
(4) Internal resistance method
The internal resistance measurement method uses alternating current of different frequencies to excite the battery, measure the internal AC resistance of the battery, and obtain the SOC estimate through the established calculation model. The battery state of charge measured by this method reflects the SOC value of the battery under a specific constant current discharge condition. Since there is no one-to-one correspondence between battery SOC and internal resistance, it is impossible to accurately model it using mathematics. Therefore, this method is rarely used in electric vehicles.
(5) Linear model method
The principle of the linear model method is a linear model established based on the changes in SOC, current, voltage and the SOC value at the previous time point. This model is suitable for situations where low current and SOC change slowly. For measurement errors and erroneous initial conditions, Has high robustness. The linear model can theoretically be applied to various types of batteries and at different aging stages, but it is currently only used in lead-acid batteries. Since the relationship between changing SOC and current and voltage is not universal, it cannot be used in other batteries. The applicability and estimation effect of variable current conditions need to be further studied.
(6) Kalman filtering method
The Kalman filter method is based on the ampere-hour integral method. The main idea of the Kalman filter method is to make the optimal estimate of the state of the dynamic system in the sense of minimum variance. This method is applied to battery SOC estimation. The battery is regarded as a power system, and the state of charge is an internal state of the system. The essence of this algorithm is that it can make optimal estimates of the state of complex dynamic systems based on the minimum mean square error principle. The nonlinear dynamic system will be linearized into the state space model of the system in the Kalman filter method. In actual application, the system updates the state variables that need to be obtained based on the estimated value at the previous moment and the observed value at the current moment, following the The “prediction-actual measurement-correction” mode eliminates random deviations and interference in the system.
Since the Kalman filter method can not only correct the initial system error, but also effectively suppress the system noise, it has significant application value in the SOC estimation of electric vehicle power batteries with very complex operating conditions.
However, this method also has two flaws: First, the accuracy of the Kalman filter method in estimating SOC depends largely on the accuracy of the battery model. The operating characteristics of the power battery itself are highly nonlinear. After the Kalman filter method, There will inevitably be errors after linearization processing. If the model is not established accurately enough, the estimated results may not be reliable. Secondly, the algorithm involved in this method is very complex, requires a huge amount of calculation, requires a long calculation period, and requires A microcontroller that requires high computing power.
(7) Neural network method
The neural network method is a new algorithm that simulates the human brain and its neurons to process nonlinear systems. It does not require in-depth study of the internal structure of the battery. It only needs to extract a large number of input and output samples from the target battery in advance that match its working characteristics, and By inputting this into a system built using this method, the running SOC value can be obtained. The post-processing of this method is relatively simple. It can not only effectively avoid the error caused by linearizing the battery model in the Kalman filter method, but also obtain the dynamic parameters of the battery in real time. However, the initial workload of the neural network method is relatively large, and a large amount of comprehensive target sample data needs to be extracted to train the system. The input training data and training methods will affect the estimation accuracy of SOC to a large extent. In addition, under the complex influence of factors such as battery temperature, self-discharge rate and battery aging, the accuracy of estimating the SOC value of the same group of batteries will be greatly reduced if this method is used for a long time. Therefore, this method is rare in SOC estimation work of power batteries.
(8) Other methods
In recent years, various SOC estimation methods have emerged in endlessly, such as support vector regression method, fuzzy logic algorithm, “offline calculation, online table lookup” fuzzy control method, analysis method, etc.、
Summarize
The methods used to estimate SOC in actual electric vehicles are all based on the traditional method, that is, adding some correction factors to the ampere-hour integral. The disadvantage is that there is a large error in the SOC estimation result, which is currently used in batteries. The SOC estimation technology of the management system is not yet very mature. Although there are many types of battery SOC estimation methods, various methods have certain flaws and are difficult to meet the requirements of real-time online and high-precision SOC estimation. In the future research on the SOC estimation method, Electric Zhijia believes that it will be improved from the following four aspects. First, through a large number of experiments, a rich database will be established to make the SOC estimation well-founded and well-documented; secondly, rely on hardware technology to improve the measurement accuracy of current, voltage, etc., and ensure the accuracy of the basic data used to estimate SOC; third, introduce an accurate battery model to more truly represent the dynamic characteristics of the battery during use; finally, comprehensively integrate various The algorithm uses its strengths to offset its weaknesses and introduces different correction methods to different stages of SOC to minimize errors in different states and improve its estimation accuracy.