Battery Degradation Evaluation Using Dynamic Weighting

Resolve Bottlenecks,
Find Innovative Solutions
Generate 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

VSEngineering 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

Engineering Contradiction:
Improvedegradation evaluation accuracyVSAvoiddata accumulation time
Core Design Contradiction:
ReliabilityVSLoss of time

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.

Inventive Principle:
Principle #10Preliminary action

2Reliability

If only internal resistance data is used for degradation evaluation, then the system is simple, but evaluation accuracy is poor

Engineering Contradiction:
Improvedegradation evaluation accuracyVSAvoidmeasurement system complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

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.

Inventive Principle:
Principle #5Merging (Combining)

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.

Inventive Principle:
Principle #6Universality (Multi-functionality)

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

Engineering Contradiction:
Improvedegradation evaluation accuracy with limited dataVSAvoidshort-term degradation precision
Core Design Contradiction:
ReliabilityVSMeasurement precision

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.

Inventive Principle:
Principle #15Dynamics

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.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP3862765B1Battery degradation evaluation system, battery degradation evaluation method, and battery degradation evaluation program
Publication Date: 2023.06.21 TOYOTA JIDOSHA KK
  • EP3862765B1 patent drawingFigure 1
  • EP3862765B1 patent drawingFigure 2
  • EP3862765B1 patent drawingFigure 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.