Catalyst Temperature Estimation Using Machine Learning Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing catalyst temperature estimation devices for internal combustion engines require extensive man-hours for adapting mapping data, especially when using ambient temperature and excess fuel amount variables, which increases computational complexity and reduces accuracy in transient states.
Innovation Solution
A catalyst temperature estimation device that utilizes machine learning to generate mapping data, incorporating variables such as ambient temperature, excess fuel amount, fluid energy, and previous cycle values, to improve estimation accuracy while reducing adaptation time, and includes features like multiple mappings for different scenarios and regions within the catalyst.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a large amount of mapping data is used to estimate catalyst temperature in transient state, then estimation accuracy is improved, but the number of man-hours for adapting the mapping data increases
Solution Approach 1:
The patent applies preliminary action by pre-calculating and storing mapping data for various operating conditions before actual use. The ECU stores pre-prepared mapping data that covers different ambient temperatures, excess air ratios, and catalyst temperatures, eliminating the need for real-time adaptation during vehicle operation. This allows the system to maintain high estimation accuracy while reducing adaptation time to minimal initial setup.
2Measurement precision
If ambient temperature variable and excess amount variable are included in mapping data, then estimation accuracy in transient state is improved, but the complexity of mapping data structure increases
Solution Approach 1:
The patent segments the mapping data structure into distinct dimensional layers: ambient temperature dimensions, excess air ratio dimensions, catalyst temperature dimensions, and correction value dimensions. Each variable is organized in separate axes of a multi-dimensional map, allowing the system to handle complex relationships between variables while maintaining a structured and manageable data format that the ECU can efficiently process.
3Measurement precision
If mapping data is adapted to reflect transient temperature changes, then estimation accuracy is improved, but the computational complexity increases
Solution Approach 1:
The patent uses copying by creating a pre-computed mapping data structure that replicates complex transient temperature relationships in advance. Instead of performing complex computational models in real-time, the system copies the essential temperature characteristics into a lookup table format that can be quickly queried and interpolated, significantly reducing computational complexity while maintaining accuracy.
Data Source
AI summary
A catalyst temperature estimation device that estimates a temperature of a catalyst provided in an exhaust passage of an internal combustion engine includes a storage device and processing circuitry. The storage device stores mapping data that specifies a mapping that uses multiple input variables to output an estimation value of the temperature of the catalyst. The multiple input variables include at least one variable of an ambient temperature variable or an excess amount variable. The multiple input variables further include a fluid energy variable, which is a state variable related to energy of fluid flowing into the catalyst, and a previous cycle value of the estimation value of the temperature of the catalyst. The processing circuitry is configured to execute an acquisition process, a temperature calculation process, and an operation process. The mapping data includes data that is learned through machine learning.