Battery Capacity Estimation Using Kalman Filter
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing battery capacity estimation methods produce noisy and inaccurate results due to dependency on historical voltage measurement values, leading to lag in estimation and noise in measurements, which is not suitable for real-time tracking of battery degradation in applications like electric and hybrid-electric vehicles.
Innovation Solution
A method that monitors battery voltage, current, and temperature to estimate voltage-based and current-based states of charge using Coulomb counting, with uncertainty quantification and Kalman filter processing to generate an updated overall battery capacity estimate, reducing lag and noise by recursively updating values and accounting for sensitivity to noise factors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If historical voltage measurement values are used for battery capacity estimation, then the estimation process is simplified, but the results become noisy and inaccurate with lag
Solution Approach 1:
The patent introduces a Kalman filter as an intermediary processing layer between the voltage measurements and the final capacity estimation. This filter mediates the relationship by recursively processing measurements and predictions, eliminating noise and lag while maintaining computational feasibility. The Kalman filter acts as a mathematical mediator that transforms raw measurements into accurate estimates without requiring complex hardware modifications.
2Productivity
If real-time tracking of battery degradation is implemented, then accurate capacity monitoring is achieved, but noise and lag in measurements increase
Solution Approach 1:
The patent implements a feedback mechanism through the Kalman filter that continuously compares predicted capacity values with actual measurements and adjusts the estimation accordingly. This feedback loop enables real-time tracking of battery degradation while filtering out noise and correcting lag, thereby maintaining high measurement reliability. The recursive nature of the Kalman filter ensures that each new measurement informs the next estimation, creating a self-correcting system.
3Ease of operation
If voltage-based state of charge determination is used, then the measurement process is straightforward, but uncertainty and noise in the data increase
Solution Approach 1:
The patent transforms the raw voltage measurement parameter into a more reliable state of charge estimate by changing its processing method. Instead of using voltage directly, the system applies the Kalman filter to recursively process voltage-based measurements, thereby converting noisy voltage data into accurate state of charge information. This parameter transformation maintains the simplicity of voltage measurement while eliminating its associated noise and uncertainty.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach reduces lag and noise in battery capacity estimation, allowing for accurate tracking of capacity degradation over time, improving the reliability of battery state estimates and enabling better decision-making for battery management and maintenance.
Implementation Method 1
processing the voltage-based state of charge and the predicted battery state of charge using a Kalman filter to generate an updated overall battery capacity estimate
Implementation Method 2
determining an integrated current value through Coulomb counting based upon the data from the sensor
Data Source
AI summary
A method for battery capacity estimation is provided. The method includes, within a computerized processor, monitoring a sensor operable to gather data regarding a battery, determining a voltage-based state of charge for the battery based upon the data from the sensor, determining a capacity degradation value for the battery based upon the data from the sensor, determining an integrated current value through Coulomb counting based upon the data from the sensor, determining a predicted battery state of charge for the battery based upon the capacity degradation value and the integrated current value, processing the voltage-based state of charge and the predicted battery state of charge using a Kalman filter to generate an updated overall battery capacity estimate, and using the updated overall battery capacity estimate to control management of the battery.