Battery Cell Formation Data for ML-Based Performance Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional battery configuration methods are costly, cumbersome, and inefficient, limiting battery lifetime and performance, especially for lithium-ion batteries which require high energy density and stability.
Innovation Solution
A system and method utilizing machine learning models to predict battery performance by measuring various parameters before and during cell formation, including impedance, open circuit voltage, and coulombic efficiency, to optimize silicon-dominant anode and cathode configurations, and employing lamination and direct coating processes to enhance electrode performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If conventional battery configuration methods are used, then manufacturing process is established, but production cost increases and production time extends
Solution Approach 1:
The patent applies preliminary action by measuring multiple parameters (impedance, open circuit voltage, coulombic efficiency) during the formation process before final battery configuration is determined. This allows performance prediction to occur in advance, enabling optimized configuration decisions without extending production time.
Solution Approach 2:
The patent replaces conventional trial-and-error mechanical configuration methods with machine learning-based prediction systems. The ML model analyzes measured parameters and predicts battery performance, substituting empirical configuration approaches with data-driven optimization that reduces both cost and time.
2Ease of manufacture
If conventional battery configuration methods are used, then manufacturing process is established, but battery lifetime is limited
Solution Approach 1:
The patent implements feedback by using machine learning models that analyze measured parameters (impedance, open circuit voltage, coulombic efficiency) and provide predictions about battery performance and lifetime. This feedback loop enables optimization of battery configuration based on actual formation process data, improving reliability without compromising manufacturability.
Solution Approach 2:
The patent applies parameter changes by monitoring and analyzing multiple electrical parameters during formation (impedance, open circuit voltage, coulombic efficiency) and using these parameter variations to predict and optimize battery lifetime. The ML model identifies optimal parameter ranges that maximize battery reliability.
3Measurement precision
If machine learning models are used to predict battery performance, then prediction accuracy improves, but measurement and data processing complexity increases
Solution Approach 1:
The patent applies universality by developing machine learning models that can predict multiple battery performance metrics (cycle life, capacity, impedance) from a single set of formation process measurements. This multi-functional approach improves prediction accuracy across different performance dimensions without proportionally increasing system complexity.
Solution Approach 2:
The patent implements self-service by using the battery's own formation process data (impedance, open circuit voltage, coulombic efficiency measurements taken during manufacturing) to train and apply machine learning models that predict its future performance. The system uses the battery's inherent characteristics without requiring external test equipment or additional complexity.
4Quantity of substance
If silicon-dominant anodes are used to increase energy density, then energy density improves, but manufacturing complexity and cost increase
Solution Approach 1:
The patent applies preliminary action by measuring key parameters (impedance, open circuit voltage, coulombic efficiency) during the formation process of silicon-dominant anodes before final configuration decisions are made. This early measurement approach enables optimization of silicon anode configurations without adding manufacturing steps or complexity.
Solution Approach 2:
The patent applies parameter changes by monitoring electrical parameters during formation and using machine learning to identify optimal configurations for silicon-dominant anodes that maximize energy density. The ML model analyzes parameter variations to determine best practices for silicon anode manufacturing without increasing process complexity.
Data Source
AI summary
Methods and systems are provided for key predictors and machine learning for configuring cell performance. One or more parameters relating to operation of a cell may be measured, via a measurement apparatus, with the cell including a cathode, a separator, and a silicon-dominant anode, and cell performance may be managed, based on the one or more parameters, with the managing including assessing the cell performance using a machine learning model. The cell may be within a battery pack that includes a plurality of cells, each of which including a cathode, a separator, and a silicon-dominant anode. One or more of the plurality of cells from the battery pack in response to a determination, based on the assessing, of a different performance of the one or more of the plurality of cells. The battery pack may be in an electric vehicle.