Today we talk about the State of Charge (SOC) arithmetic method in battery management systems (BMS). SOC is a very important parameter index in electronics and electrical industry, which is involved in Smart phone, E-bike, energy storage system, EV, etc. In the lithium iron phosphate (LFP) battery system, because of its special electrochemical characteristics, the charge-discharge curve is not linear, that makes the accurate calculation of SOC become particularly difficult.
Here are some common approach used in BMS to estimate the SOC of a Lithium Iron Phosphate (LFP) battery. Those method relies on measuring the battery’s voltage and using a lookup table or mathematical model to correlate the voltage with the battery’s SOC.
Ampere-Hour Integration: This method estimates SOC by integrating the current flowing in and out of the battery over time, similar to coulomb counting. However, it doesn’t considers the battery’s capacity variations due to aging and temperature effects. We should use a more sophisticated algorithm to adjust the SOC calculation based on these factors.
Voltage-Based: This method estimates SOC based on the battery’s open-circuit voltage (OCV) characteristics. It uses a lookup table or mathematical model to correlate the battery voltage with its SOC. However, the accuracy of this method can be affected by factors like battery aging and load conditions especially high power load.
Internal resistance method: In the process of charging and discharging, some parameters in the battery impedance change with the change of SOC. According to this characteristic, SOC can be estimated by internal resistance. According to the method of obtaining internal resistance, the internal resistance method can be divided into DC internal resistance method and AC internal resistance method. This method is suitable for auxiliary calculation
Kalman Filtering: This is an algorithm that uses the linear system state equation to optimally estimate the system state through the input and output observation data of the system. This method combines multiple measurements, such as voltage, current, and temperature, using a Kalman filter algorithm to estimate SOC. It takes into account the uncertainties and noise in the measurements to provide a more accurate SOC estimation.
Neural network method: When estimating the battery SOC, relevant data such as temperature, current, and voltage can be used as the input of the neural network, and the SOC value can be used as the output of the neural network. Through sample data training, a more accurate SOC value can be obtained. SOC estimation using neural network can achieve accurate estimation of nonlinear system through enough samples without knowing the internal structure of the battery.
It’s important to note that BMS solution company may use a combination of these methods or other proprietary algorithms to estimate SOC accurately based on the specific battery chemistry and characteristics. For example, in energy storage application, many company will use ampere-hour integration as a basic, and lookup OCV table for calibration. Some will insert more factors(voltage, IR, temperature, time..) and setup a modeling calculation. In EV application, companies will choose Kalman Filtering or Neural network method for higher accuracy.
