Battery Parameter Estimation Using Frequency-Based Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing battery management systems in hybrid electric vehicles face challenges in accurately estimating battery parameters, such as internal resistance and impedance, which affects power capability prediction and longevity, often requiring extensive offline testing and being sensitive to noise and environmental conditions.
Innovation Solution
Implementing a battery management method that filters terminal voltage data into high-frequency and low-frequency content, using a Randles Circuit Model and Extended Kalman Filter to estimate internal resistance and impedance, allowing for real-time power capability prediction without additional hardware or increased complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional battery management systems estimate battery parameters without frequency-based filtering, then the estimation process is simpler, but the accuracy and noise sensitivity of parameter estimation deteriorates
Solution Approach 1:
The patent segments the battery parameter estimation process into two distinct frequency-based pathways: low-frequency analysis for internal resistance estimation and high-frequency analysis for internal impedance estimation. This segmentation allows each pathway to focus on specific frequency ranges, improving measurement precision while keeping each individual pathway relatively simple.
Solution Approach 2:
The patent introduces a frequency dimension to the parameter estimation process by filtering terminal voltage data into high-frequency and low-frequency components. This dimensional transformation enables the system to extract different parameter information from different frequency ranges, significantly improving estimation accuracy without substantially increasing system complexity.
2Measurement precision
If offline testing is used to determine battery parameters, then comprehensive parameter data can be obtained, but the time consumption and system response capability deteriorates
Solution Approach 1:
The patent performs preliminary frequency-based filtering and parameter extraction by continuously processing terminal voltage data in real-time. The controller applies low-pass and high-pass filters to separate frequency components, enabling real-time parameter estimation without requiring subsequent offline processing.
Solution Approach 2:
The patent replaces traditional offline mechanical/electrical testing methods with a computational approach using the Extended Kalman Filter algorithm. This substitution allows the system to estimate battery parameters in real-time based on normal operating voltage and current data, eliminating the need for separate offline testing procedures.
3Reliability
If frequency-based filtering is applied to terminal voltage data, then noise sensitivity is reduced, but the computational complexity increases
Solution Approach 1:
The patent implements dynamic frequency-based filtering where the system adaptively processes terminal voltage data through low-pass and high-pass filters based on real-time operating conditions. The Extended Kalman Filter dynamically updates parameter estimates using the filtered signals, allowing the system to maintain reliability by reducing noise sensitivity while managing computational complexity through adaptive processing.
Solution Approach 2:
The patent introduces frequency filtering as an intermediary step between raw terminal voltage measurement and parameter estimation. By inserting low-pass and high-pass filters as intermediaries, the system separates noise components from useful signal components, significantly reducing noise sensitivity. The computational complexity is managed by using standard filtering algorithms that are computationally efficient.
Data Source
AI summary
A hybrid powertrain system includes a battery and at least one controller. The at least one controller is configured to determine instantaneous battery power limits during operation of the system using filtered battery voltage signals and current input signals. The at least one controller is further configured to separate medium-to-high frequency dynamics of the measured battery voltage. The filtering process, in certain examples, is realized using a low pass filter or a high pass filter. The at least one controller is further configured to correlate the medium-to-high frequency loads to estimate battery parameters and determine battery dynamics using, in one example, an Extended Kalman Filter.