Regression Coefficient Correction for Battery Thermal Network Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods struggle to estimate coefficients with high precision due to multicollinearity issues in time-series data, leading to inaccurate modeling of physical phenomena, particularly in thermal networks of battery modules.
Innovation Solution
An information processing device employs sparse estimation technology and correction values based on prior knowledge to estimate coefficients, using a regression equation generation module and an estimation module to select valid variables and correct coefficients, thereby improving precision.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional regression methods are used to estimate coefficients from time-series data, then the modeling process is simple, but the precision of coefficient estimation deteriorates due to multicollinearity issues
Solution Approach 1:
The system performs preliminary actions by identifying and removing collinear variables before coefficient estimation, and by pre-calibrating correction values using known relationships between variables. This prevents multicollinearity issues from degrading estimation precision while maintaining modeling simplicity.
Solution Approach 2:
Correction values are introduced as intermediaries between the observed time-series data and the final coefficient estimates. These correction values compensate for multicollinearity effects and improve estimation precision without requiring complex alternative modeling approaches.
2Adaptability or versatility
If all input variables are included in the regression model, then the model captures more physical phenomena, but the precision of individual coefficient estimates deteriorates due to multicollinearity
Solution Approach 1:
The set of input variables is segmented into collinear and non-collinear groups. Variables within collinear groups are processed specially through correction mechanisms, while non-collinear variables are included directly in the regression model, preserving both model coverage and estimation precision.
Solution Approach 2:
Different processing approaches are applied to different variables based on their collinearity characteristics. Variables with high collinearity receive correction value adjustments, while independent variables use standard regression estimation, optimizing precision for each variable's specific properties.
3Reliability
If correction values based on prior knowledge are applied, then the precision of physical phenomenon modeling improves, but the complexity of the modeling process increases
Solution Approach 1:
Correction values based on prior physical knowledge are calculated and applied in advance to the regression coefficients. This preliminary correction step improves modeling accuracy by incorporating domain knowledge without requiring complex iterative optimization during the main modeling process.
Solution Approach 2:
The system uses feedback from known physical relationships to adjust coefficient estimates. Correction values are derived from established physical laws and relationships, providing feedback that refines the raw regression coefficients into more accurate physical parameter estimates.
Data Source
AI summary
An information processing device includes a regression equation generation module and an estimation module. The regression equation generation module generates a regression equation that includes a plurality of input variables and a plurality of coefficients respectively corresponding to the plurality of input variables to obtain one or more output variables. The estimation module estimates a plurality of coefficients by using one or more correction values each of which corrects each of one or more target coefficients included in the plurality of coefficients, where the correction values is determined based on knowledge regarding the relationship between the input variable corresponding to the target coefficient and the output variable.