Battery Cell Diagnosis Using Voltage Variance Trends
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing battery diagnosis systems face challenges in accurately detecting abnormal battery cells due to weak signal strength, irregular signal periods, and noise in sensing signals, leading to uncertainties and reduced accuracy in identifying potential safety hazards.
Innovation Solution
A battery diagnosis apparatus and method that calculates voltage variance and accumulative variance based on Z-scores, using average and standard deviation to standardize voltage data, thereby enhancing the detection of abnormal battery cells.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If voltage sensing signals are used to detect abnormal battery cells, then the detection process can be implemented, but the signal strength is weak and contains noise, reducing detection accuracy
Solution Approach 1:
The patent transforms the raw voltage sensing signals into voltage variance parameters by calculating the difference between voltage values at different times. This parameter transformation converts the weak, noisy voltage signals into variance data that better reflects abnormal battery cell characteristics, thereby improving detection accuracy while maintaining reliability.
Solution Approach 2:
The patent introduces voltage variance as an intermediary parameter between the raw voltage sensing signals and the final abnormal cell detection. This intermediary variable processes the weak signals through mathematical operations (calculating differences and variances) to produce more reliable detection data, effectively mediating between the noisy sensor output and the required accurate detection.
2Device complexity
If only sensing signals are used for diagnosis, then the system remains simple, but the irregular signal period and noise degrade detection accuracy
Solution Approach 1:
The patent performs preliminary processing on the sensing signals by calculating voltage differences and variances before the actual detection occurs. This preliminary action transforms the raw signals into processed variance data that is more suitable for detection, preparing the data in advance to improve accuracy without adding complex hardware, only computational steps.
Solution Approach 2:
The patent replaces direct mechanical/electrical signal analysis with mathematical processing of voltage data. Instead of relying on the physical characteristics of the sensing signals, the system uses computational methods (calculating variances and comparing against thresholds) to achieve accurate detection, substituting physical signal analysis with mathematical processing.
3Measurement precision
If voltage variance calculation is performed for each battery cell, then detection accuracy improves, but calculation time and processing complexity increase
Solution Approach 1:
The patent calculates voltage variance for all battery cells (excessive action) to ensure comprehensive detection coverage, then uses efficient comparison methods (variance against threshold) that minimize processing time. By applying the variance calculation to all cells uniformly and using simple threshold comparison rather than complex analysis, the system achieves high detection accuracy while controlling processing time through scalable computation.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A battery diagnosis apparatus according to an embodiment disclosed herein includes a sensor configured to obtain voltage data of each of battery cells and a processor configured to calculate a voltage variance dV that is a difference between voltage values at two arbitrary points in time for each of the battery cells, based on the voltage data, calculate an accumulative variance based on the voltage variance, and diagnose whether the battery cells are abnormal, based on the accumulative variance.