Battery SOH Regression for Vehicle Deterioration Prediction

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing methods for predicting battery deterioration in electrically-driven vehicles face challenges due to large current changes in short times, making it difficult to obtain a deterioration coefficient for each category, and the computing process is complex.

Innovation Solution

A computing system that acquires travel data, specifies State of Health (SOH) of batteries, generates deterioration regression curves, and predicts remaining life using a regression function based on average travel distance or discharge amount, adjusting for changes in travel conditions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a method of obtaining deterioration coefficient for each category is used, then prediction accuracy can be improved, but the computing process becomes complicated

Engineering Contradiction:

Improvedeterioration prediction accuracyVSAvoidcomputing process complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the deterioration prediction process into two parts: (1) obtaining individual deterioration coefficients for each battery through curve regression on time-series SOH data, and (2) establishing a separate regression function relating deterioration coefficients to travel conditions. This segmentation avoids the complexity of categorizing all possible operating conditions while maintaining prediction accuracy through individualized battery modeling.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent changes the approach from using fixed category-based deterioration coefficients to using dynamically calculated individual coefficients through curve regression. By fitting deterioration curves to actual time-series SOH data for each battery and extracting individual coefficients, the system adapts to each battery's unique degradation pattern without requiring complex pre-defined categories.

Inventive Principle:

Principle #35Parameter changes

2Measurement precision

If deterioration coefficients are obtained for each category, then prediction precision is improved, but the method becomes difficult to apply to batteries with large current changes

Engineering Contradiction:

Improvedeterioration coefficient accuracyVSAvoidapplicability to varying operating conditions

Core Design Contradiction:

Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The patent transitions from static category-based deterioration coefficients to dynamic individual coefficients derived from actual time-series data through curve regression. This dynamic approach allows the system to adapt to large current changes and varying operating conditions by continuously modeling each battery's actual degradation behavior rather than forcing it into fixed categories.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

Each battery serves itself by generating its own deterioration coefficient through curve regression on its own time-series SOH data. This self-service approach eliminates the need for external categorization and makes the system adaptable to any battery regardless of its specific operating conditions, including those with large current changes.

Inventive Principle:

Principle #25Self-service

3Measurement precision

If individual deterioration coefficients are obtained for each battery through curve regression, then prediction accuracy for specific batteries is improved, but data requirements increase

Engineering Contradiction:

Improveindividual battery prediction accuracyVSAvoiddata quantity required

Core Design Contradiction:

Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent applies partial action by obtaining deterioration coefficients only for the specific batteries that need prediction, rather than pre-calculating coefficients for all possible batteries and conditions. The curve regression is performed on actual available time-series data for each target battery, using only the necessary amount of data required for individual prediction accuracy.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS12558993B2Computing system, battery deterioration predicting method, and battery deterioration predicting program

Publication Date: 2026.02.24 PANASONIC INTELLECTUAL PROPERTY MANAGEMENT CO LTD

AI summary

Deterioration regression curve generation unit generates a deterioration regression curve of each battery by performing curve regression on a plurality of SOHs specified in time series for each battery. Coefficient regression function generation unit generates a regression function of a deterioration coefficient using an average travel distance or an average discharge amount per unit period of a plurality of electrically-driven mobile units as an independent variable and using a deterioration coefficient of the deterioration regression curve of each of the plurality of the batteries as a dependent variable. Deterioration prediction unit specifies the average travel distance or the average discharge amount per unit period in accordance with the received change in the travel conditions, applies the average travel distance or the average discharge amount per unit period to the regression function of the deterioration coefficient to specify a deterioration coefficient after the change in the travel conditions, and uses the deterioration coefficient to change the deterioration regression curve of battery mounted in electrically-driven mobile unit.