Battery SOC Estimation Using Pre-Depolarization Regression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating the state of health (SOH) of secondary batteries, particularly in vehicles with frequent charging and discharging cycles, face challenges in accurately estimating SOH before depolarization due to prolonged depolarization times, leading to insufficient estimation accuracy.
Innovation Solution
A state-of-charge estimation method that uses regression analysis based on battery voltage data after charging or discharging to correct estimated SOC values, classifying correlations between Δ-pseudo-SOC values and their rate of change to accurately estimate post-depolarization SOC, even before depolarization occurs, thereby improving estimation accuracy and frequency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If depolarization time is extended to improve SOH estimation accuracy, then measurement precision improves, but productivity decreases due to reduced computation frequency
Solution Approach 1:
The patent performs SOC estimation using pseudo-SOC values and regression analysis before depolarization is complete, rather than waiting for full depolarization. This preliminary action allows the system to generate estimation results earlier in the process, increasing computation frequency while maintaining accuracy through the regression-based correction method.
Solution Approach 2:
The patent changes the approach from waiting for depolarization completion to using real-time pseudo-SOC values combined with regression analysis. By transforming the estimation methodology to work with intermediate depolarization states through mathematical modeling, the system achieves both high computation frequency and maintained accuracy.
2Productivity
If regression analysis is performed using pseudo-SOC values before depolarization, then productivity increases through higher computation frequency, but measurement precision may deteriorate due to polarization effects
Solution Approach 1:
The patent employs regression analysis that uses the relationship between pseudo-SOC values and actual SOC values to continuously correct estimation errors. This feedback mechanism processes multiple data points and adjusts the estimation model, compensating for polarization effects and maintaining high measurement precision even when computing before full depolarization.
Solution Approach 2:
The patent combines multiple elements - pseudo-SOC values, regression analysis, and correction algorithms - to create a composite estimation method. This composite approach integrates several computational techniques to achieve both high computation frequency and maintained accuracy, overcoming the limitations of individual methods.
3Measurement precision
If depolarization time is prolonged in vehicles with frequent charging/discharging cycles, then measurement precision improves, but loss of time increases
Solution Approach 1:
The patent performs SOC estimation before depolarization is complete by using pseudo-SOC values and regression analysis. This preliminary estimation approach eliminates the need to wait for full depolarization, significantly reducing the time loss while maintaining estimation accuracy through the regression-based correction method.
Solution Approach 2:
The patent skips the traditional waiting period for depolarization completion by implementing real-time estimation using pseudo-SOC values. The regression analysis allows the system to rush through the estimation process before full depolarization occurs, reducing time loss without sacrificing measurement precision.
Data Source
AI summary
Provided is a state-of-charge estimation method that includes: obtaining a plurality of pseudo-SOC values before depolarization, based on a cell voltage after a battery 11 is continuously charged or discharged and the charging or discharging is stopped; setting an initial value for an estimated SOC value that is estimated as a post-depolarization SOC; obtaining a Δ-pseudo-SOC value that is a difference between each of the pseudo-SOC values and the estimated SOC value; obtaining a correlation between a square of the Δ-pseudo-SOC value and a rate of change of the Δ-pseudo-SOC value; classifying the correlation into a first region in which there is non-linear change and a second region in which there is linear change, and obtaining a regression line for the second region; and using the regression line to correct the estimated SOC value such that the estimated SOC value approaches a post-depolarization true SOC value.