Collision Avoidance Target Selection for Turning Vehicles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing collision avoidance assist systems face challenges in efficiently selecting targets in a front-and-lateral region, leading to increased processing loads and reduced stability of collision avoidance control, especially when the number of detected targets is high, as they struggle to accurately identify potential collision threats under uniformly set selection conditions.
Innovation Solution
A collision avoidance assist apparatus that includes a front target information acquisition device, a front-and-lateral target information acquisition device, and a control unit capable of adjusting selection conditions based on the vehicle's turning state, selecting targets with higher collision possibility by differentiating between straight and turning scenarios, and prioritizing targets based on speed and position to reduce processing load and improve selection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If the number of detected targets increases, then the coverage of detection region increases, but the processing load increases and stability of collision avoidance control decreases
Solution Approach 1:
The patent extracts and processes only the most critical target information needed for collision avoidance control. By selectively extracting essential parameters (position, speed, collision risk) from the full set of detected target data, the system reduces processing load while maintaining detection coverage across multiple sensors.
Solution Approach 2:
The patent applies different processing quality levels to different targets based on their collision risk. High-priority targets (those with high collision possibility) receive detailed processing, while low-priority targets receive simplified processing or are excluded, thereby optimizing the balance between processing load and control stability.
2Ease of operation
If uniform selection conditions are applied to all targets, then the selection process is simplified, but targets with collision possibility may not be appropriately selected
Solution Approach 1:
The patent dynamically adjusts selection conditions based on the vehicle's operating state (straight motion vs. turning). When the vehicle is turning, the selection conditions are modified to prioritize targets in the turning direction, ensuring that collision risks specific to turning maneuvers are appropriately identified while maintaining operational simplicity.
Solution Approach 2:
The patent changes selection parameters (such as speed thresholds, position weights, and collision risk coefficients) according to the vehicle's motion state. This allows the system to adapt to different driving scenarios without requiring completely different selection algorithms, balancing simplicity with accuracy.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Collision avoidance assist control is appropriately executed by appropriately selecting detected targets while processing load is reduced. A collision avoidance assist apparatus includes: a front-and-lateral target information acquisition device (13L, 13R) configured to acquire front-and-lateral target information; a vehicle information acquisition device (14, 15, 16) configured to acquire vehicle information including a vehicle speed and at least one of a yaw rate or a steering input value; and a control unit (10) configured to execute, as collision avoidance assist control, at least one of warning control or automatic braking control when a target satisfies a collision condition that is satisfied when a collision risk is determined to exist based on the front-and-lateral target information. The control unit selects targets that satisfy a predetermined selection condition from targets included in the front-and-lateral target information, determines whether the collision condition is satisfied for each of the selected targets, determines, when this determination is to be made, whether an own vehicle is turning based on the vehicle information, and changes the predetermined selection condition between a case in which the own vehicle is not turning and a case in which the own vehicle is turning.