Battery Capacity Estimation Using Dynamic Multi-Stage Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional coulometers lack accuracy and stability in estimating battery capacity due to jitter errors and low current errors, and they often use constant noise parameters that fail to account for varying noise conditions.
Innovation Solution
A dynamic filter algorithm is employed to estimate battery capacity by sampling voltages across a sensor resistor, applying average, median, and low-pass filtering to remove white noise, impulse noise, and adjust filtering thresholds based on data scope, using first or second-order low-pass filters as needed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional coulometers output data without filter processing, then the calculation is simple and fast, but the data contains jitter errors and low current errors resulting in poor accuracy
Solution Approach 1:
The filtering process is divided into three distinct segments: average filtering to remove white noise, median filtering to remove impulse noise, and low-pass filtering to smooth remaining fluctuations. Each segment addresses specific types of errors independently, achieving comprehensive noise removal while maintaining clear functional separation in the processing pipeline.
Solution Approach 2:
The patent implements dynamic adjustment of filtering parameters based on real-time noise characteristics. The system adaptively changes filtering strength according to detected noise levels, transitioning between different filtering intensities to maintain optimal performance across varying operating conditions while preventing over-filtering of valid signal changes.
2Reliability
If conventional coulometers use constant noise parameter for filtering, then the filtering process is simple, but it fails to account for varying noise conditions affecting stability
Solution Approach 1:
The system continuously monitors the filtered data to detect remaining noise characteristics and feeds this information back to adjust filtering parameters. This closed-loop approach allows the system to automatically adapt to changing noise conditions, maintaining optimal filtering performance and stability without manual intervention or complex external control systems.
Solution Approach 2:
The patent dynamically changes filtering parameters including noise thresholds and filtering coefficients based on detected operating conditions and noise levels. By adjusting these parameters in real-time according to actual noise characteristics rather than using fixed constants, the system maintains high reliability across diverse operating scenarios while keeping the adaptation logic relatively simple.
3Measurement precision
If multiple filtering stages are applied to remove different noise types, then the accuracy improves, but the processing time and computational load increase
Solution Approach 1:
Different filtering operations are segmented and applied in sequence, each targeting specific noise types. Average filtering handles white noise efficiently, median filtering quickly removes impulse noise, and low-pass filtering smooths remaining variations. This segmented approach processes only relevant noise components at each stage rather than applying heavy computation uniformly, reducing overall processing time while maintaining accuracy.
Solution Approach 2:
The system applies filtering selectively based on detected noise characteristics rather than always applying maximum filtering strength. When noise levels are low or data is stable, filtering intensity is reduced or skipped, avoiding unnecessary computational overhead. This partial action approach maintains accuracy when needed while minimizing processing time during stable operating conditions.
Data Source
AI summary
A method for estimating battery capacity of a battery, includes: sampling a series of voltages across a sensor resistor coupled to the battery when a current flows through the sensor resistor under conditions of default time and default temperature, and recording the voltages as sample voltages; calculating average filtering voltages by filtering white noise in the sample voltages; calculating median filtering voltages by removing impulse noise in the average filtering voltages; detecting data scope of the median filtering voltages in the data characteristic filed; calculating low-pass filtering voltages by using a low-pass filter formula according to the median filtering voltages domain in the data characteristic filed; and estimating the battery capacity of the battery according to the low-pass filtering voltages.