Battery Degradation Prediction via Optimal Exponent Parameter Transformation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing battery degradation prediction methods, such as those using neural networks or gradient-boosted decision trees, are complex and difficult to interpret, and struggle to accurately model the nonlinear degradation trend of secondary batteries, making it challenging to construct a battery model with high prediction accuracy while maintaining a simple structure.
Innovation Solution
A battery model construction method that treats powers of usage history parameters, such as current, voltage, and temperature, as explanatory variables, and searches for an optimal exponent to generate input parameters for a linear regression model, allowing for accurate prediction of battery degradation while retaining a simple structure.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If methods based on neural network or GBDT are used to construct battery model, then prediction accuracy is improved, but model complexity increases and structure becomes complex
Solution Approach 1:
The patent transforms the input parameters by raising them to optimal exponents (e.g., current^0.5, voltage^0.5, temperature^0.5) before feeding into a simple linear regression model. This parameter transformation enables the model to capture nonlinear degradation trends while maintaining the simplicity of linear regression, thus achieving high prediction accuracy without increasing model complexity.
2Measurement precision
If methods based on neural network or GBDT are used to construct battery model, then prediction accuracy is improved, but model construction time increases
Solution Approach 1:
By transforming parameters through exponentiation and using a linear regression model with predetermined optimal exponents, the patent dramatically reduces model construction time compared to training complex neural networks or GBDT models, while still achieving accurate prediction of battery degradation.
3Device complexity
If simple linear regression model is used to construct battery model, then model structure remains simple, but prediction accuracy deteriorates due to inability to capture nonlinear degradation trend
Solution Approach 1:
The patent applies parameter transformation by raising usage history parameters to optimal exponents before inputting them into the linear regression model. This allows the simple linear model to effectively capture nonlinear degradation trends, achieving both structural simplicity and high prediction accuracy.
Data Source
AI summary
A battery model construction method is a method for constructing a battery model that treats powers of a plurality of usage history parameters defined on a basis of time series data about a current, a voltage, and a temperature of a battery as explanatory variables and treats a predicted SOH value as an objective variable, the method including: an acquiring step ST1 for acquiring time series data of usage history parameters and measured SOH values; an exponentiating step ST2 for raising the usage history parameters by a prescribed exponent to thereby generate time series data of input parameters; a training step ST4 for training the battery model by using the time series data of the input parameters and the measured SOH values as training data; and a searching step ST8 for searching for an optimal exponent by repeatedly performing the steps ST2 and ST4 while varying the exponent.