Battery SoH Estimation via Kalman Filter Convergence
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining the state of health (SoH) of rechargeable batteries, particularly Li-ion batteries, face challenges due to varying battery types and chemistries, and the unreliability of voltage-based SoC estimation, especially for batteries with low SoH.
Innovation Solution
A method that classifies the battery's SoH as good, medium, or poor before charging, using an extended Kalman filter to repeatedly measure and predict terminal voltages during charging, converging to a reliable SoC value, which is then used to calculate the SoH based on charge supplied and rated capacity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If voltage method is used to estimate initial SoC, then estimation can be obtained quickly, but reliability is poor particularly for Li ion batteries and batteries with low SoH
Solution Approach 1:
The system performs preliminary classification of battery SoH into categories (good, medium, poor) before charging begins. This preliminary action allows the system to select appropriate estimation strategies and adjust parameters in advance, improving both speed and accuracy of subsequent SoC estimation without requiring full discharge cycles
Solution Approach 2:
The system implements iterative feedback during charging by repeatedly measuring terminal voltage, comparing with predicted values from the extended Kalman filter, and adjusting the SoC estimate. The process continues until convergence is achieved (difference below threshold), ensuring reliable SoC determination while maintaining efficiency through adaptive iteration rather than fixed prolonged measurement
2Measurement precision
If coulomb counting is used during charge cycle, then full charge capacity can be obtained, but accurate SoH determination requires battery to be fully discharged prior to charging which is not always practical
Solution Approach 1:
The system performs preliminary SoH classification before charging begins using the extended Kalman filter with initial voltage measurements. This allows the system to establish a baseline SoH estimate without requiring full discharge, making the process operationally convenient while maintaining measurement precision through the filter's ability to handle partial charge states
Solution Approach 2:
The system uses the charging process itself to determine SoH by measuring voltage during charging and using the extended Kalman filter to extract SoH information. The battery's own charging behavior provides the data needed for SoH determination, eliminating the need for separate discharge-charge cycles and making the process self-sufficient
3Reliability
If extended Kalman filter is used with repeated measurements during charging, then reliable converged SoC value is obtained, but charging time is extended
Solution Approach 1:
The system performs preliminary SoH classification before charging begins, which allows the extended Kalman filter to start with informed initial parameters. This preliminary preparation reduces the number of iterations needed during charging to achieve convergence, thereby maintaining reliability while reducing the time penalty
Solution Approach 2:
The system performs repeated voltage measurements and filter updates during charging, but stops once convergence is achieved (when difference between predicted and measured voltage is below threshold). This partial action approach performs just enough measurements to achieve reliable convergence without unnecessary additional measurements that would extend charging time
Data Source
AI summary
A rechargeable battery is charged and, at the same time, steps are taken to assess its state of health (SoH). The SoH is first broadly classified, which is only a coarse estimate. Next, the state of charge (SoC) is estimated, and then the SoC is repeatedly measured and predicted during charging until its value converges to a reliable value. The SoH is a state in the Kalman filter used for determining the SoC, and the SoH is also refined through the filter process. The SoH is reported when or after the SoC has converged, and is reported before charging is completed.