Battery Cell Fault Diagnosis Using Moving-Average Voltage Deviations
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional battery management systems struggle to accurately diagnose abnormal battery cells due to noise interference and external factors, leading to misdiagnosis and potential safety risks.
Innovation Solution
A battery management apparatus and method that utilizes long and short moving averages of battery cell voltages to calculate deviations, distinguishing between normal and abnormal behaviors by setting threshold values and time criteria, thereby reducing noise interference and misdiagnosis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If voltage deviation diagnosis is performed using individual battery cell voltage compared to average voltage, then abnormal battery cells can be identified, but noise interference causes vulnerability and reduces measurement precision
Solution Approach 1:
The patent applies preliminary action by calculating moving averages of battery cell voltages before performing deviation diagnosis. The controller calculates a first moving average and a second moving average of voltages from multiple battery cells, then uses these pre-processed values as the basis for deviation calculation. This preliminary smoothing of voltage data removes noise interference before the actual diagnosis, improving measurement precision without requiring additional hardware.
2Extent of automation
If a threshold criterion is set for diagnosing abnormal battery cells, then diagnosis can be automated, but the threshold cannot be adjusted below a certain level due to noise vulnerability
Solution Approach 1:
The patent uses preliminary action by pre-calculating moving averages of battery voltages before threshold comparison. The controller computes the first moving average from individual cell voltages and the second moving average from multiple cells, then calculates deviation based on these smoothed values. This allows the threshold to be set at appropriate levels without being overwhelmed by noise, enabling reliable automatic diagnosis.
Solution Approach 2:
The patent introduces an intermediary element - the moving average calculation - between the raw voltage measurements and the threshold comparison. By using the first and second moving averages as intermediaries, the system filters out noise while preserving actual abnormal voltage deviations, allowing the threshold criterion to function effectively for automatic diagnosis.
3Device complexity
If conventional voltage deviation methods are used, then simple diagnosis can be performed, but micro-disconnection abnormalities cannot be detected
Solution Approach 1:
The patent applies preliminary action by calculating moving averages before deviation analysis. The controller computes the first moving average of individual battery cell voltages and the second moving average of multiple battery cell voltages, then uses these pre-processed values to calculate deviation. This preliminary smoothing enhances the ability to detect subtle micro-disconnection abnormalities while maintaining diagnostic simplicity.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A battery management apparatus according to an embodiment disclosed herein includes a voltage measurement unit configured to measure a voltage of each of a plurality of battery cells and a controller configured to calculate a first deviation, which is a deviation between a long moving average and a short moving average of a battery cell voltage of each of the plurality of battery cells, calculate a second deviation, which is a deviation between a long moving average and a short moving average of an average voltage of the plurality of battery cells, and calculate a first diagnosis deviation between the first deviation and the second deviation for each of the plurality of battery cells, diagnose each of the plurality of battery cells based on whether a final diagnosis deviation related to the first diagnosis deviation of each of the plurality of battery cells is at least a first threshold value and based on a time during which the final diagnosis deviation is maintained as being at least a second threshold value.