Abstract
In the field of electric vehicles, the performance of battery management systems (BMS) and the effective life of batteries are key considerations. To improve the service life of the battery pack, it is necessary to periodically balance the batteries. Traditionally, battery balancing mainly relies on passive balancing technology, which converts excess energy into thermal energy to achieve balance between batteries. However, this method not only causes heat management problems, but also reduces the overall efficiency of the battery pack.
This article proposes an innovative active balancing strategy that uses the Kalman filter algorithm to optimize the efficiency of BMS, effectively addressing the shortcomings of passive balancing technology. The core goal is to build a system that can uniformly manage battery charging and discharging, thereby extending the battery's lifespan. The system has designed an active balancing circuit that utilizes the Kalman filter algorithm to accurately estimate the state of each battery and calculate the optimal charging and discharging current based on it, in order to achieve efficient balancing between batteries.
Research background, plan and results
1. Research background and motivation
The development background of electric vehicles and the importance of battery management systems: The global attention to environmental pollution and fuel price increases caused by automobile exhaust emissions highlights the necessity of deploying electric vehicles. The innovation of Battery Management Systems (BMS) has made electric vehicles a powerful candidate for future transportation, but BMS still has many areas for improvement to enhance efficiency and reliability.
Key elements and challenges of battery management system
The importance of SOC and SOH estimation: Accurately estimating the state of charge (SOC) and state of health (SOH) of a battery is crucial for the reliable and efficient operation of BMS. SOC measures the available capacity of a battery relative to its fully charged state, while SOH indicates the degree of battery aging, reflecting the difference in energy storage capacity between the current fully charged state and the manufacturing state.
Challenge and Balancing Requirements in Battery Pack Design: Designing a safe and energy-efficient battery pack is extremely challenging, as it requires hundreds of volts of DC voltage and hundreds of kilowatts of power, consisting of a large number of batteries in series and parallel. However, due to manufacturing defects and aging, the parameters of the battery do not match, which reduces the effective capacity of the battery pack. Therefore, BMS and external balancing circuits are needed to fully utilize the energy of each battery. Battery balancing circuits are divided into passive and active balancing. Passive balancing converts battery energy into thermal energy through shunt resistors to prevent overcharging, while active balancing uses DC/DC converters or other power transfer methods to directly transfer energy between batteries. Implementing an active balancing circuit can improve the safety, durability, charging and discharging performance, and energy utilization efficiency of battery packs.
2. Propose a plan
Overall architecture and working principle: The proposed scheme architecture (see Figure 1) includes SOC estimation (using extended Kalman filter algorithm), BMS controller, and active equalization circuit. The controller senses the SOC of each battery and sends signals to the active balancing circuit to transfer charge from high SOC batteries to low SOC batteries, ultimately balancing the charge of each battery in the battery pack.
SOC estimation method
Extended Kalman Filter Algorithm Process: SOC estimation adopts the Extended Kalman Filter algorithm, which is a repetitive process that considers noise and errors in the instrument and estimation. Firstly, determine the various attributes and their dependencies of the battery, and use a lumped parameter model to design the equivalent circuit model of the battery.
By analyzing the circuit using Kirchhoff's Voltage Law (KVL), the terminal voltage equation is derived:
Applying Kirchhoff's current law (KCL) to derive the RC branch equation, based on the relationship between battery SOC and circuit current:
Establish a continuous time state space model, then convert it to a discrete-time state space model (using a closed form discretization formula to process the correlation matrix and vectors), and finally apply the Kalman filter algorithm for SOC estimation (including state equations and measurement equations, noise is an independent zero mean Gaussian process, calculation includes time update and measurement update steps).
The principle of buck boost converter: A buck boost converter is a DC-DC converter, and the output voltage can be lower or higher than the input voltage. When the switch is turned on (MOSFET closed, diode off), the inductor stores energy; When the switch is turned off (MOSFET is turned off, diode is turned on), the inductor releases energy to the load, and the output voltage increases. Its working mode is divided into two situations.
Working mechanism of active balancing circuit: In the active balancing circuit, the controller senses the SOC imbalance between batteries, determines the direction of charge transfer, and sends PWM signals to control the switch. If the controller detects that the top battery N needs to transfer energy to the bottom battery N-1, it sends a signal to the switch S2N. After the inductor stores energy to the maximum value, the switch is closed, the inductor voltage is reversed, and the diode D_N-1 is forward biased. Energy is transferred to the battery N-1 through the diode, and vice versa.
3. Simulation results
SOC estimation algorithm validation: In Matlab, the SOC estimated by the extended Kalman filter algorithm is consistent with the actual SOC over time curve, indicating that the algorithm has been successfully used to estimate battery SOC.
Evaluation of Active Balancing Circuit Effectiveness: Using a Matlab simulation model of an active balancing circuit with a buck boost converter, the initial SOC of the upper and lower batteries were set to 23% and 20%, respectively. After simulation, the final balanced SOC of the upper and lower batteries were 21.39% and 21.4%, respectively, which were close to the initial average SOC and successfully achieved charge balancing. By changing parameters such as inductance value, cycle, and duty cycle, it was found that there is a trade-off between balancing time and final balancing SOC. For example, when the inductance value decreases, the cycle increases, or the duty cycle changes, the balance time and final SOC will change accordingly. Specifically, the smaller the inductance value, the larger the cycle, and the duty cycle changes within a certain range, the shorter the balance time, but the final SOC will also be affected to some extent.
| L(inductance) in H | Time taken to balance in sec | Final SOC (%) |
| 1 | 423 | 21.45 |
| 0.5 | 228 | 21.4 |
| 0.1 | 80 | 21.02 |
| 0.01 | 39 | 20.16 |
| 0.001 | 34 | 21.5 |
| Period (s) | Time taken to balance in sec | Final SOC (%) |
| 1 | 329 | 21.44 |
| 1.5 | 228 | 21.4 |
| 2 | 187 | 21.36 |
| 2.5 | 143 | 21.34 |
| Duty cycle (%) | Time taken to balance in sec | Final SOC(%) |
| 30 | 594 | 21.45 |
| 40 | 340 | 21.43 |
| 50 | 228 | 21.4 |
| 60 | 72 | 21.2 |
| 70 | 51 | 20.93 |
Summary
Research on Active Balancing Technology: This article focuses on the active balancing technology of single battery charge level balance in battery packs. During the completion of the project, an active balancing circuit was designed and circuit simulation was conducted to obtain the expected results.
Selection of SOC estimation methods: Multiple single battery SOC estimation methods were studied, and the extended Kalman filter method was ultimately adopted due to its accuracy in estimating nonlinear parameters.
Research verification: Overall, the project has successfully demonstrated the effectiveness of active balancing in improving battery performance and reducing safety risks. Through simulation, the active balancing circuit can achieve a balance state close to the average SOC for batteries with different initial SOC, indicating that it can effectively improve battery performance and reduce safety hazards that may be caused by battery imbalance.
The importance of considering specific requirements: The study also emphasizes the need to carefully consider the specific requirements of battery systems and applications when determining the most suitable active balancing system. Different battery systems (such as battery packs composed of different types of batteries and battery usage requirements in different application scenarios) may have different requirements for active balancing systems, such as different emphasis on balancing speed, balancing accuracy, energy loss, etc. Therefore, the most suitable active balancing scheme needs to be selected according to the actual situation to achieve optimal performance and safety.