Battery Voltage Drop Clustering for Defect Screening
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for screening low-voltage defective batteries fail to reliably distinguish between good and defective products due to manufacturing deviations, leading to potential over-detection or non-detection of defective batteries.
Innovation Solution
A method that includes pre-filtering to remove outlier data, clustering based on similar characteristics, correcting the dispersion of voltage drop measurements according to temperature or period, and screening low-voltage defective batteries using anomaly detection algorithms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a specific threshold is used for screening low-voltage defective batteries, then the screening process is simple, but the detection reliability decreases due to manufacturing deviations
Solution Approach 1:
The patent applies preliminary action by performing pre-filtering to remove outlier data before the main screening process. This removes extreme values that would otherwise distort the reference value calculation, allowing the subsequent threshold-based screening to be both simple and reliable. The outlier removal is performed in advance, so the main screening can still use simple threshold comparison without being affected by manufacturing deviations.
Solution Approach 2:
The patent implements feedback by iteratively calculating the reference value from filtered data and using it to identify outliers, then repeating the process. This feedback loop allows the system to automatically adjust and refine the reference value based on the actual data distribution, improving detection reliability while maintaining the simplicity of threshold-based screening in the final step.
2Productivity
If conventional threshold-based screening is used, then the process is fast and simple, but manufacturing deviations cause over-detection or non-detection of defective batteries
Solution Approach 1:
The patent performs preliminary outlier removal before the main screening to eliminate data points that would cause over-detection or non-detection. This pre-processing step is computationally efficient and allows the subsequent threshold-based screening to proceed quickly while achieving higher accuracy by excluding distorted data points from the reference value calculation.
Solution Approach 2:
The patent replaces the simple mechanical threshold comparison with a more sophisticated statistical approach that uses standard deviation and iterative outlier removal. This substitution maintains productivity by using efficient computational algorithms while dramatically improving measurement precision by accounting for manufacturing deviations through statistical analysis.
3Device complexity
If no pre-filtering is performed, then the screening process is straightforward, but the dispersion of voltage drop data increases due to outliers, blurring the boundary between good and defective products
Solution Approach 1:
The patent applies preliminary action by removing outliers before calculating the reference value and performing screening. This pre-filtering step reduces the dispersion of voltage drop data by eliminating extreme values, thereby sharpening the boundary between good and defective products. The process remains relatively simple because it uses straightforward statistical calculations and iterative refinement.
Solution Approach 2:
The patent uses feedback by iteratively recalculating the reference value and standard deviation after removing outliers, then identifying new outliers based on the updated statistics. This iterative feedback process progressively reduces data dispersion and refines the boundary between good and defective products, achieving high manufacturing precision without excessive complexity.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Disclosed herein relates to a method for screening a low-voltage defective battery including: (a) a pre-filtering step of collecting data and removing secondary batteries having an outlier data in real time during an activation process of multiple secondary batteries; (b) a clustering step of clustering based on similar characteristics or records for multiple pre-filtered secondary batteries; (c) a correction step of measuring the amount of voltage drop (ΔOCV) for each cluster and correcting a dispersion of the amount of voltage drop (ΔOCV) according to a temperature or a period; and (d) a screening step of screening low-voltage defective batteries based on a corrected dispersion of the amount of voltage drop.