Gray Box Model for Vehicle Component Warming-Up Behavior
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current statistical models for modeling and monitoring the warming-up behavior of vehicle components are inaccurate, complex, and often applied too late in the development process, leading to unjustified error messages and high expenditures.
Innovation Solution
A method involving the creation of a gray box model using a white box model and a black box model, where input variables are detected and simulated to optimize the training of the black box model, allowing for more accurate and early modeling of warming-up behavior, with the black box model based on a neural network and the white box model providing physical boundary conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If statistical reference curves are used to model warming-up behavior, then security with respect to legal requirements is guaranteed, but accuracy and reliability of the model deteriorates
Solution Approach 1:
The patent segments the modeling approach into three distinct components: a white box model for physical boundary conditions, a black box model for statistical patterns, and a gray box model as their integrated combination. This segmentation allows each component to serve its specific purpose - the white box ensures physical accuracy, the black box captures statistical variations, and together they provide both legal security and modeling accuracy.
Solution Approach 2:
The patent creates a composite modeling system by combining different modeling approaches (white box, black box, and gray box models) into a unified framework. This composite approach integrates the strengths of physical-based modeling with statistical modeling, achieving both accuracy and reliability that neither approach could achieve alone.
2Reliability
If statistical models are used for warming-up behavior, then legal requirements are met, but the modeling process becomes complex and time-consuming
Solution Approach 1:
By dividing the modeling process into three distinct stages (white box model creation, black box model training, and gray box model integration), the patent reduces the complexity of the overall process. Each stage can be performed independently using specialized techniques, making the complex task more manageable and efficient.
Solution Approach 2:
The white box model is created first to establish physical boundary conditions and simulate input variables before the black box model is trained. This preliminary action prepares the data and framework needed for subsequent statistical modeling, reducing the complexity of the overall process by breaking it down into sequential, manageable steps.
3Reliability
If statistical models are used, then legal security is ensured, but the modeling can only be performed late in the development process
Solution Approach 1:
The white box model can be created and validated early in the development process before final statistical models are needed. This preliminary modeling allows for early detection of issues and optimization of physical boundary conditions, enabling the complete modeling process to be performed earlier in development rather than only at the end.
Solution Approach 2:
By separating the modeling into independent components (white box, black box, and gray box), the patent enables parallel development and validation. The white box model can be developed and validated independently in earlier stages, while the black box model is trained subsequently, reducing the overall development time compared to sequential approaches.
4Ease of manufacture
If inaccurate models are used in OBD systems, then development is simplified, but user experience deteriorates due to unjustified error messages
Solution Approach 1:
The composite modeling approach combines the simplicity of statistical models with the accuracy of physical-based models. The white box model provides physically accurate boundary conditions that prevent unjustified error messages, while the black box model maintains the statistical simplicity needed for easy development. Together, they deliver both ease of manufacture and improved user experience.
Data Source
AI summary
A method for modeling a warming-up behavior of a vehicle component includes detecting a plurality of input variables and obtaining warming-up curves during a plurality of warming-up processes of the vehicle component. The method also includes creating a white box model based on items of information with respect to at least one of the plurality of input variables, and training and validating a black box model based on items of information comprising at least one output variable of the white box model and at least one of the warming-up curves. The method further includes modeling a warming-up behavior of the vehicle component using a gray box model formed of the white box model and the black box model.