Battery SOC Clustering via Representative Cell Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing battery state of charge (SOC) estimation methods, such as coulomb counting and electric circuit models, face reduced accuracy in degraded or extreme temperature conditions, and require significant computation, necessitating a more efficient approach.
Innovation Solution
An apparatus and method that clusters battery cells into groups based on similarity in sensing data, selects a representative cell for each group, and estimates the SOC using either an electric or electro-chemical model, allowing for faster and more accurate SOC determination of the battery pack.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If electro-chemical model is used for SOC estimation, then estimation accuracy is improved, but computation time increases
Solution Approach 1:
The battery pack is divided into multiple cell groups based on clustering cells with similar characteristics (voltage, temperature, current). Instead of modeling every cell individually, only representative cells from each group are modeled, segmenting the computation task into manageable portions that maintain accuracy while reducing overall computational burden.
Solution Approach 2:
SOC values from representative cells are copied and applied to estimate SOC of other cells within the same group. This copying approach leverages the similarity among cells in the same group, allowing accurate SOC estimation for multiple cells through a single representative cell's model output.
2Measurement precision
If all cells are modeled individually, then SOC estimation accuracy is improved, but device complexity increases
Solution Approach 1:
The battery management system is segmented into multiple cell groups rather than treating all cells uniformly. This segmentation reduces the number of independent models needed while maintaining comprehensive monitoring through representative cells from each segment.
Solution Approach 2:
Representative cells serve multiple functions: they represent their own individual cell characteristics and simultaneously represent all cells within their group. This multi-functionality reduces the total number of models required while maintaining accurate estimation across the entire battery pack.
3Measurement precision
If representative cell selection is performed frequently, then SOC estimation accuracy is improved, but computation speed decreases
Solution Approach 1:
Representative cell selection and group updates are performed periodically at predetermined intervals rather than continuously. This periodic action maintains accurate SOC estimation by updating models at appropriate intervals while avoiding unnecessary computations between updates, thus preserving computation speed.
Data Source
AI summary
Battery state of charge (SOC) determination apparatuses and methods are disclosed, where the battery SOC determination apparatus includes a grouper configured to cluster the cells in the battery pack into the groups based on similarity of the sensing data among the cells, a representative cell selector configured to select the representative cell for the each group, a first SOC estimator configured to estimate the SOC of the representative cell of the each group based on the battery model, and a second SOC estimator configured to determine the SOC of the battery pack based on the SOC of the representative cell of the each group.