Battery Degradation Evaluation Using Dynamic Weighting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing battery degradation evaluation systems are ineffective in accurately assessing battery degradation until sufficient data is collected, and their evaluation accuracy is poor when internal resistances are not considered.
Innovation Solution
A battery degradation evaluation system that uses a computation model to derive short-term, medium-term, and long-term degradation probabilities based on acquired state quantities, with higher weighting for combined probabilities when data is scarce and short-term probabilities when data is abundant, and employs machine learning with training data from different vehicles to improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If battery degradation evaluation is performed using accumulated internal resistance data, then evaluation can be conducted, but evaluation accuracy is poor when data is insufficient
Solution Approach 1:
The system performs preliminary actions by acquiring and utilizing multiple types of state quantities (voltage, temperature, current, power) before sufficient internal resistance data accumulates. The computation model is trained in advance with diverse state quantity data to enable accurate degradation probability derivation even when internal resistance data alone is insufficient for reliable evaluation.
2Reliability
If only internal resistance data is used for degradation evaluation, then the system is simple, but evaluation accuracy is poor
Solution Approach 1:
The system merges multiple state quantity measurements (voltage, temperature, current, power) with internal resistance data in the computation model. By combining these diverse data sources, the system achieves accurate degradation evaluation without requiring any single measurement system to be overly complex, as each individual sensor remains relatively simple.
Solution Approach 2:
The computation model serves multiple functions: it derives degradation probabilities from various state quantities, adapts to different data availability conditions, and provides comprehensive degradation assessment. This multi-functionality allows the system to maintain accuracy across different operating conditions without requiring separate specialized systems for each function.
3Reliability
If high weighting is given to long-term degradation probability when data is scarce, then accurate evaluation is possible with limited data, but short-term degradation precision may be reduced
Solution Approach 1:
The system dynamically adjusts the weighting between long-term and short-term degradation probabilities based on data availability. When internal resistance data is scarce, higher weighting is given to long-term probability derived from diverse state quantities. As more data accumulates, the weighting shifts to prioritize short-term precision. This dynamic adaptation resolves the contradiction by optimizing the evaluation approach according to current data conditions.
Solution Approach 2:
The system changes parameters (weighting factors) based on data availability conditions. The computation model adjusts the relative importance of different degradation probability components depending on whether sufficient internal resistance data has been accumulated, thereby optimizing evaluation accuracy for the current data state without permanently sacrificing short-term precision capability.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A battery degradation evaluation system (10) comprises an acquisition section (22, 30), a derivation section (42), and an evaluation section (44). The acquisition section (22, 30) acquires state quantities of a battery (20) mounted at a vehicle (14). The derivation section (42), based on the state quantity acquired by the acquisition section and a pre-specified computation model, derives each of a short-term degradation probability of the battery (20) degrading in a pre-specified short period, a medium-term degradation probability of the battery (20) degrading in a medium period that is longer than the short period, and a long-term degradation probability of the battery (20) degrading in a long period that is longer than the medium period. The evaluation section (44) evaluates degradation of the battery (20) based on derivation results from the derivation section (42), sets a higher weighting for a combined degradation probability for the short period and the medium period when a number of state quantities is smaller, the combined degradation probability being calculated from the long-term degradation probability, and sets a higher weighting for the short-term degradation probability when the number of state quantities is larger. The state quantities being used as data for learning of the computation model.