Battery Management System SOC Estimation Using Extended Kalman Filter
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing battery state of charge (SOC) estimation methods, such as those using the extended Kalman filter, face inaccuracies due to measurement errors and the difficulty in simulating nonlinear battery characteristics, especially when the SOC is close to full discharge.
Innovation Solution
A battery management system that utilizes an extended Kalman filter to estimate SOC by determining open circuit voltage (OCV) information based on comparisons with a reference range and adjusts the reference range's maximum value according to battery degradation, thereby reducing the impact of nonlinear characteristics on SOC estimation accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If ampere counting is used to estimate SOC, then the estimation is simple to implement, but measurement errors and external noise cause discrepancy between estimated SOC and actual SOC
Solution Approach 1:
The patent combines ampere counting with equivalent circuit model and extended Kalman filter to create a hybrid estimation approach. The ampere counting provides cumulative charge information while the equivalent circuit model and EKF compensate for measurement errors and nonlinear effects, achieving both simplicity and accuracy.
Solution Approach 2:
The extended Kalman filter implements feedback by continuously comparing the estimated SOC with actual battery behavior and adjusting the estimation accordingly. This feedback mechanism corrects cumulative errors in ampere counting and adapts to changing battery conditions.
2Adaptability or versatility
If equivalent circuit model is used to simulate battery characteristics, then the model can represent battery behavior, but it is very difficult to sufficiently simulate the nonlinear characteristic associated with rapid OCV change near full discharge
Solution Approach 1:
The patent employs a dynamic equivalent circuit model with time-varying parameters that adapt to different battery states. The model parameters are updated in real-time based on operating conditions, enabling accurate simulation of nonlinear characteristics during rapid OCV changes near full discharge.
Solution Approach 2:
The patent changes model parameters dynamically based on battery state, particularly adjusting parameters that govern OCV behavior in the near-full-discharge region. This allows the model to accurately represent rapid OCV changes that occur in this critical range.
3Reliability
If extended Kalman filter is used to combine ampere counting and equivalent circuit model, then the drawbacks of each method are compensated, but when nonlinear characteristic is very strong, error still occurs in SOC estimation
Solution Approach 1:
The patent segments the SOC estimation process into different regions, with special handling for the near-full-discharge region where nonlinear effects are strongest. By dividing the estimation domain and applying region-specific correction strategies, the patent maintains accuracy even under strong nonlinear conditions.
Solution Approach 2:
The patent introduces an intermediary correction mechanism that specifically addresses the nonlinear OCV behavior. This intermediary layer processes the extended Kalman filter output and applies additional corrections tailored to strong nonlinear conditions, reducing estimation errors in critical regions.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Provided are a battery management system, a battery management method, a battery pack and an electric vehicle. The battery management system includes a sensing unit to generate battery information indicating a current, a voltage and a temperature of a battery, and a control unit. The control unit determines a temporary estimate for a SOC in a current cycle using a time update process of an extended Kalman filter based on a previous estimate indicating a SOC in a previous cycle and the battery information. The control unit determines open circuit voltage (OCV) information based on the temporary estimate. The control unit determines a definitive estimate indicating the SOC in the current cycle using a measurement update process of the extended Kalman filter based on the temporary estimate, the OCV information and the battery information.