Collision Avoidance Steering with Adjacent Lane Space Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing collision avoidance assist systems for vehicles face challenges in detecting obstacles in a wider area ahead, leading to limitations in performing autonomous steering to avoid collisions, especially when the collision avoidance space is not available within the host vehicle's lane, and require additional sensors that increase cost and computational load.
Innovation Solution
The system employs a combination of front direction camera and radar sensors to detect objects within a specific area ahead, determining the presence of a collision avoidance space in an adjacent lane and ensuring safety by assessing surrounding conditions to permit autonomous steering control, thereby avoiding collisions without the need for extensive additional sensors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If additional sensors (e.g., front side direction radar sensors) are arranged at the right front corner and left front corner of the vehicle body to detect objects in a wider area, then the detection area is enlarged and object recognition accuracy is improved, but the system cost and development time increase significantly
Solution Approach 1:
The front direction radar sensor is designed to perform multiple functions: detecting objects in the front direction and providing sufficient information for collision avoidance decisions. By making this single sensor multi-functional, the system avoids the need for multiple specialized sensors (front side direction radar sensors at corners), thereby reducing system complexity and cost while maintaining adequate detection capability for the collision avoidance space determination
Solution Approach 2:
The patent extracts only the essential detection function needed for collision avoidance by using a front direction radar sensor with a predetermined detection area, rather than implementing a comprehensive multi-sensor system. This extraction approach focuses on detecting objects in the critical collision avoidance space without the overhead of additional sensors, reducing system complexity while addressing the core safety requirement
2Adaptability or versatility
If additional sensors are added to detect objects in a wider area ahead, then the operational scenarios for collision avoidance are increased, but the calculation load and memory capacity requirements increase
Solution Approach 1:
The front direction radar sensor is designed to perform multiple functions: detecting objects in the front direction and providing sufficient information for collision avoidance decisions. By making this single sensor multi-functional, the system avoids the need for multiple specialized sensors (front side direction radar sensors at corners), thereby reducing system complexity and cost while maintaining adequate detection capability for the collision avoidance space determination
Solution Approach 2:
The patent implements a detection area that extends to the diagonally forward left and right directions, which is wider than the minimum necessary for straight-ahead collision avoidance. This partial expansion of detection coverage enables the system to detect objects in adjacent lanes without requiring full 360-degree coverage, thereby increasing adaptability for lane-changing collision avoidance while keeping computational requirements manageable
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This configuration allows for safer and more effective autonomous steering by determining the presence of a collision avoidance space in adjacent lanes, increasing operational scenarios for collision avoidance while maintaining a simple and cost-effective system design.
Implementation Method 1
a front direction camera sensor configured to take a picture of a front direction camera detection area to obtain front direction camera sensor information
Implementation Method 2
a front direction radar sensor configured to obtain front direction radar sensor information that is information on an object that is present in a front direction radar detection area
Data Source
AI summary
A driving assist ECU determines that a current situation is a specific situation where it is predicted that there is no object that is about to enter an adjacent lane from an area outside of a host vehicle road on which a host vehicle is traveling, when a road-side object is detected at a part around an edge of the adjacent lane, and/or when a white line painted to define the adjacent lane is detected at the part around the edge of the adjacent lane and no object near the detected white line is detected. The driving assist ECU does not perform a steering control for avoiding a collision, the steering control for letting the vehicle enter the adjacent lane, when it is not determined that the current situation is the specific situation.