Vehicle Control Device Prioritizing Unconverged Learning Regions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle control systems face challenges in achieving optimal performance due to biased learning control corrections based on driver-specific usage situations and traveling environments, leading to difficulties in improving vehicle performance across all learning regions.
Innovation Solution
A control device and method that prioritizes automatic driving control to select and adjust vehicle states and routes to unconverged learning regions, promoting learning control convergence by changing driving states and routes when necessary, and establishing a learning permission condition to facilitate early learning control.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If learning control is performed for each learning region according to driver-specific usage situations, then vehicle performance can be improved according to individual driving styles, but the learning control becomes biased toward specific learning regions and fails to improve performance across all regions
Solution Approach 1:
The system monitors the convergence state of learning control across multiple learning regions and uses this feedback to dynamically adjust travel route selection. When certain learning regions are identified as unconverged, the system preferentially selects routes that pass through these regions, creating a closed-loop control mechanism that ensures balanced learning across all regions while maintaining adaptability to driver behavior.
2Measurement precision
If the system preferentially selects travel routes through unconverged learning regions, then learning control convergence is accelerated across all regions, but this may require changing current travel routes from driver-preferred paths
Solution Approach 1:
The system performs preliminary assessment of learning control convergence status across all learning regions before route selection. By identifying unconverged regions in advance, the system can proactively plan routes that facilitate learning in these regions, rather than reactively adjusting routes after convergence issues are detected. This preliminary action allows the system to balance learning objectives with route planning.
3Loss of time
If learning control is performed without considering convergence status of different regions, then the control system remains simple, but it takes longer to achieve appropriate traveling state across all learning regions
Solution Approach 1:
The system segments the learning control process into multiple learning regions based on driving conditions and vehicle states. By dividing the overall learning task into region-specific segments, the system can track convergence status independently for each region and apply targeted route selection strategies. This segmentation approach manages complexity by organizing learning control into manageable units while accelerating overall convergence.
Data Source
AI summary
While automatic driving control is being performed, traveling in a driving state of a vehicle corresponding to an unconverged region (including an unperformed region and a performed region) is preferentially selected between the traveling in the driving state of the vehicle corresponding to the unconverged region, and traveling in the driving state of the vehicle corresponding to a converged region. As such, learning control that corrects an amount of operation associated with control of the vehicle is performed more easily throughout the entire learning regions regardless of a usage state of the vehicle by a driver. Therefore, it is possible to achieve an appropriate traveling state at an early stage by the learning control that corrects the amount of operation associated with control of the vehicle.