Battery Anode Potential Estimation for Fast-Charging Plating Control

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The challenge of fast charging lithium-ion batteries is the accelerated degradation due to lithium plating, which compromises battery lifetime and efficiency, and existing methods for estimating anode potential are costly, complex, or interfere with battery operations.

Innovation Solution

A machine-learning framework using readily available battery management system data to estimate anode potential, comprising two ML systems: one for SoH indicators and another for real-time plating potential estimation, controlling charging current to prevent lithium plating.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If fast charging is applied to lithium-ion batteries, then charging speed is improved, but battery degradation accelerates due to lithium plating

Engineering Contradiction:
Improvecharging speedVSAvoidbattery lifetime
Core Design Contradiction:
SpeedVSReliability

Solution Approach 1:

The system dynamically adjusts charging parameters (current, voltage) based on real-time anode potential estimation and battery state (SoH, temperature) to prevent lithium plating while maintaining fast charging speeds. The charging current is modulated to keep anode potential above the plating threshold.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The system implements closed-loop feedback by continuously monitoring battery state (SoH, temperature, current, voltage) and using this information to adjust charging parameters in real-time, preventing lithium plating while enabling fast charging.

Inventive Principle:
Principle #23Feedback

2Measurement precision

If sensors are installed to directly measure anode potential, then measurement precision is improved, but device complexity increases

Engineering Contradiction:
Improveanode potential measurementVSAvoidsensor installation complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system uses machine learning models as intermediaries to estimate anode potential indirectly from easily measurable quantities (current, voltage, temperature, SoH) without requiring direct potential sensors. The ML models translate readily available battery management system data into accurate anode potential estimates.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system replaces physical measurement sensors with computational models (machine learning algorithms) that calculate anode potential from electrical and thermal measurements, eliminating the need for complex sensor installations while maintaining measurement accuracy.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

3Reliability

If comprehensive machine learning framework is implemented for real-time anode potential estimation, then reliability is improved, but computational requirements increase

Engineering Contradiction:
Improveanode potential estimation accuracyVSAvoidcomputational energy consumption
Core Design Contradiction:
ReliabilityVSUse of energy by moving object

Solution Approach 1:

The system performs preliminary training of machine learning models offline using extensive battery data, so that during real-time operation, only lightweight inference computations are required. The computationally intensive model development is done beforehand, enabling fast real-time estimation with minimal energy consumption.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4636417A1A method for real-time estimation of battery anode potential
Publication Date: 2025.10.22 NINGBO GEELY AUTOMOBILE RES & DEV CO LTD
  • EP4636417A1 patent drawingFigure 1a~1b
  • EP4636417A1 patent drawingFigure 1c~1d
  • EP4636417A1 patent drawingFigure 1e

AI summary

A method for real-time estimation of an anode potential of a battery (20) in connection with a battery charging event, comprising: obtaining at least one quantified State of Health [SoH] indicator of the battery (20) as output from a first machine-learning system (14); and feeding the obtained at least one quantified SoH indicator, together with battery charging data associated with the charging event, as input to a second machine-learning system (15), which provides an estimate of the anode potential of the battery (20) as output.