Lane Keeping Assist System Using Virtual Rear Image for Long Vehicles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional lane keeping assist systems (LKAS) often fail to prevent lane departure from the rear portion of vehicles, especially those with greater lengths, as they rely solely on front image sensor data, leading to potential deviations.
Innovation Solution
The system includes an image sensor to capture the front of the vehicle, calculating position errors and curvature for both the front and rear portions, and determining a final steering torque control value using these calculations and vehicle speed to control steering effectively for lane keeping, considering the length of the vehicle.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If only front image sensor data is used for lane keeping control, then the system complexity is reduced and ease of operation is improved, but lane keeping reliability deteriorates for vehicles with great length
Solution Approach 1:
The vehicle is segmented into front and rear portions for independent lane keeping control. The system divides the vehicle length into multiple sections (front portion and rear portion) and applies separate steering torque control values to each section, ensuring that both front and rear portions maintain proper lane positioning even for long vehicles.
2Reliability
If steering torque control is applied to both front and rear portions of the vehicle, then lane keeping reliability is improved, but device complexity increases
Solution Approach 1:
The system uses image copying technique by generating a virtual rear image from the front image sensor data. Instead of installing separate physical sensors at the front and rear, the rear portion's lane information is obtained by processing and transforming the front image data, thereby reducing hardware complexity while maintaining dual-zone control capability.
Data Source
AI summary
A lane keeping assist system, including: a first calculation unit which calculates left and right position errors of a front portion of the vehicle, a curvature of a lane of the front portion, and a head angle formed based on the lane from a front image of the vehicle captured by an image sensor, and calculates a target yaw rate using the left and right position errors, the curvature of the lane, and the head angle formed based on the lane which are calculated; a second calculation unit which calculates a steering torque control value for driving the calculated target yaw rate using vehicle condition information including front and rear yaw rates of the vehicle sensed by a yaw rate sensor, and a steering angle and a speed of the vehicle; and a steering driving unit which controls steering of the vehicle for keeping a lane.