Adaptive Following Control Using Driver-Learned Gap Preferences
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies fail to specify appropriate timings for learning a driver's preference for vehicle-to-vehicle information during manual driving, leading to potential inaccuracies in learning.
Innovation Solution
A control device for vehicles that includes a processor and memory, which executes learning control to acquire vehicle-to-vehicle information during manual driving, reflecting the learning results in automated driving, and determines data storage based on preceding vehicle type and environmental conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If learning control is executed during manual driving without specifying timing, then learning data can be continuously collected, but the learning accuracy deteriorates due to inappropriate timing
Solution Approach 1:
The system performs preliminary assessment of learning suitability before executing learning control. The determination unit evaluates whether current driving conditions are appropriate for learning based on vehicle-to-vehicle distance, driver behavior patterns, and road conditions, only then proceeding with data collection. This prevents inappropriate learning execution and ensures high learning accuracy.
Solution Approach 2:
The system dynamically adjusts learning execution based on changing driving parameters. When vehicle-to-vehicle distance becomes too small or driver behavior indicates inattention, the system stops learning control. This parameter-based control ensures learning occurs only under suitable conditions, resolving the contradiction between continuous data collection and learning accuracy.
2Quantity of substance
If learning control is executed at all times during manual driving, then more learning data is collected, but learning quality deteriorates due to inappropriate learning conditions
Solution Approach 1:
The system extracts and separates appropriate learning opportunities from general driving situations. By using the determination unit to identify specific conditions (suitable vehicle-to-vehicle distance, appropriate driver behavior), the system extracts only high-quality learning data moments, discarding inappropriate situations. This ensures learning data quantity is maximized while maintaining high learning quality.
3Adaptability or versatility
If vehicle-to-vehicle information is stored for all preceding vehicle types, then learning coverage is comprehensive, but learning accuracy deteriorates due to vehicle type variations
Solution Approach 1:
The system applies different learning strategies for different preceding vehicle types. The determination unit identifies vehicle type and adjusts learning parameters accordingly, recognizing that learning conditions for large vehicles differ from small vehicles. This localized approach maintains comprehensive coverage while ensuring accuracy for each vehicle type category.
Data Source
AI summary
A control device for a vehicle controls a subject vehicle configured to switch between manual driving by a driver and automated driving. The control device includes a processor and a memory device. The processor is configured to execute learning control that learns vehicle-to-vehicle information being vehicle-to-vehicle time or vehicle-to-vehicle distance of the subject vehicle with respect to a preceding vehicle during traveling with the manual driving, and to reflect a learning result of the vehicle-to-vehicle information by the learning control in a control of the vehicle-to-vehicle information during the automated driving. The memory device is configured to store the vehicle-to-vehicle information as learning data during the manual driving. In the learning control, the processor determines, based on a type of the preceding vehicle, whether or not the vehicle-to-vehicle information for the preceding vehicle is caused to be stored in the memory device as the learning data.