Intersection Collision Braking Thresholds for Cross-Traffic Avoidance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing collision avoidance assistance devices face challenges in accurately predicting collisions at intersections, leading to excessive brake applications due to decreased object recognition accuracy and path prediction accuracy as the distance from the host vehicle to the object increases, resulting in inconvenience and load for the driver.
Innovation Solution
A collision avoidance assistance device that calculates the predicted paths of the host vehicle and an object, determines the time to collision, and applies a predetermined deceleration or braking force if the time to collision is equal to or less than a threshold, while adjusting the brake application determination threshold based on the predicted passing time and the time to reach the collision point.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the brake is applied in a position such that the host vehicle can stop without entering the intersection based on the time to collision, then the host vehicle can avoid collision with the object, but the distance required for the host vehicle to stop increases as the host vehicle speed increases, causing the brake to be applied in a position distant from the object and reducing object recognition accuracy and path prediction accuracy
Solution Approach 1:
The brake application determination threshold is made dynamic by adjusting it based on the predicted passing time of the object and the amount of time to reach the predicted collision point. This allows the threshold to adapt to different scenarios (orthogonal vs. non-orthogonal intersections, different object speeds), enabling the system to apply brakes at optimal distances that maintain both collision avoidance effectiveness and prediction accuracy.
Solution Approach 2:
The system changes the brake application determination threshold parameter based on the type of intersection and object characteristics. For orthogonal intersections where objects pass through quickly, a smaller threshold is used to allow later brake application. For non-orthogonal intersections where objects remain in the path longer, a larger threshold is used to ensure earlier brake application, thus maintaining prediction accuracy while achieving collision avoidance.
2Reliability
If the brake application determination threshold is set to ensure collision avoidance, then the host vehicle can stop in time, but excessive brake applications occur when the threshold is too large, leading to inconvenience and load for the driver
Solution Approach 1:
Different brake application determination thresholds are applied to different local situations: orthogonal intersections receive a smaller threshold to prevent excessive braking, while non-orthogonal intersections receive a larger threshold to ensure safety. This localized adaptation of the threshold parameter resolves the contradiction between ensuring collision avoidance and preventing excessive brake applications that inconvenience drivers.
3Speed
If the brake application determination threshold is increased to allow later brake application, then the host vehicle can maintain higher speed for longer, but the accuracy of predicting whether or not the host vehicle and the object will collide decreases
Solution Approach 1:
The system dynamically adjusts the brake application determination threshold based on the predicted passing time and intersection type. This dynamic adjustment allows the threshold to be increased only when safe (in non-orthogonal intersections with longer passing times), enabling later brake application and speed maintenance, while keeping the threshold smaller in orthogonal intersections to maintain prediction accuracy.
Data Source
AI summary
Provided is a collision avoidance assistance device with improved functionality and usability, which can suppress an excessive application or emergency application of the brakes in assisting avoidance of a collision with an object traveling in a lane orthogonal to a host vehicle (a road crossing the host vehicle). The collision avoidance assistance device calculates an amount of time to reach a predicted collision point that is an estimated amount of time until the host vehicle reaches the predicted collision point when the host vehicle is decelerated by a predetermined deceleration or braking force, and changes (increases or decreases) a brake application determination threshold based on a predicted passing time and the amount of time to reach the predicted collision point, the predicted passing time being an estimated time required for the object to pass through an intersecting region (predicted intersecting region) between a predicted path of the object (predicted target path) and a predicted path of the host vehicle (predicted host vehicle path).