Battery Cell Consistency Detection Using SOC Regression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing battery management systems struggle to accurately detect the consistency of battery packs due to the lack of sufficient historical data for capacity and self-discharge calculations, making it challenging to implement precise cumulative capacity data, especially in cloud-based systems.
Innovation Solution
A battery detection method that utilizes multiple linear regression analysis based on state of charge values of a reference battery cell and other cells, along with timestamps, to determine relative battery capacity and self-discharge without requiring accurate cumulative capacity data, thereby improving detection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If cumulative capacity data is used for consistency detection, then detection accuracy can be improved, but data acquisition difficulty increases
Solution Approach 1:
The patent creates a virtual reference battery cell that copies the expected behavior of an ideal battery cell based on linear regression models. This virtual copy allows consistency detection to proceed without requiring actual cumulative capacity data from all battery cells, resolving the contradiction between needing accurate data and the difficulty of acquiring it.
Solution Approach 2:
The patent introduces linear regression analysis as an intermediary method that bridges the gap between easily obtainable state of charge data and the harder-to-obtain cumulative capacity data. By using regression models to estimate capacity relationships, the system achieves accurate consistency detection without direct measurement of cumulative capacity.
2Measurement precision
If all battery cells are compared individually for consistency detection, then detection accuracy improves, but calculation complexity increases
Solution Approach 1:
The patent extracts one battery cell as a reference cell and removes it from the comparison group. By comparing all other cells against this single reference rather than performing pairwise comparisons among all cells, the system maintains detection accuracy while significantly reducing calculation complexity.
3Ease of operation
If state of charge values are used instead of cumulative capacity data, then data acquisition becomes easier, but traditional consistency detection methods fail
Solution Approach 1:
The patent fundamentally changes the parameter basis for consistency detection from cumulative capacity data to state of charge values combined with linear regression analysis. This parameter transformation allows the system to use easily acquired state of charge data while maintaining detection reliability through mathematical modeling that establishes capacity relationships.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present invention relates to the field of battery technologies and discloses a battery detection method, apparatus, and device, and a storage medium. The battery detection method includes: acquiring (S10) state of charge values of each battery cell in a battery system at different time points within a preset time; and performing multiple linear regression analysis (S30) based on state of charge values of a reference battery cell, the state of charge values of the battery cells, and timestamps at corresponding time points to obtain a consistency detection result of the battery cells in the battery system. In the present invention, a consistency detection result of the battery cells in the battery system is obtained by just selecting the reference battery cell from the battery cells and comparing the reference battery cell with the battery cells, without using accurate cumulative capacity data. Even without the capacity data, a relative result can be obtained through calculation, thus consistency of the battery cells is calculated, and the accuracy of battery detection is improved.