Pre-Collision Control Device Radar Virtual Image Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional pre-collision control devices face accuracy issues in identifying unnecessary objects detected by radar sensors without using image sensor information, leading to potential false pre-collision controls.
Innovation Solution
A pre-collision control device that uses millimeter wave radars to detect objects and determines whether they are unnecessary by analyzing their location and velocity changes within specific detection areas and velocity ranges, thereby reducing the likelihood of performing pre-collision controls on misdetected objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the conventional device uses image sensor information to determine whether a radar object is a virtual image object, then the identification accuracy may improve, but the system complexity increases and the image object may be incorrectly selected/extracted decreasing the accuracy
Solution Approach 1:
The invention extracts and removes the dependency on image sensor information from the radar object identification process. By using only radar sensor data and analyzing the spatial distribution and movement characteristics of radar objects, the system eliminates the need for image sensor integration, thereby reducing system complexity while maintaining identification accuracy through alternative measurement approaches
Solution Approach 2:
The radar sensor is made multi-functional by enabling it to perform both collision risk assessment and virtual image object identification using the same sensor data. The analysis of spatial distribution patterns and movement characteristics allows the radar sensor to serve dual purposes, eliminating the need for separate image sensor systems while maintaining identification capabilities
2Reliability
If the device performs pre-collision control on all detected objects, then the safety coverage is maximized, but unnecessary pre-collision controls increase causing false alarms
Solution Approach 1:
The system performs preliminary identification of virtual image objects before executing pre-collision control actions. By analyzing the spatial distribution and movement characteristics of detected objects in advance, the system identifies and filters out virtual image objects, ensuring that pre-collision controls are only applied to real objects, thereby eliminating false alarms while maintaining comprehensive safety coverage
Solution Approach 2:
The system uses feedback from continuous analysis of radar object spatial distribution and movement characteristics to dynamically adjust pre-collision control decisions. By monitoring whether objects exhibit patterns consistent with virtual image objects, the system provides feedback to the control execution stage, enabling real-time filtering of false targets while maintaining responsiveness to genuine collision risks
3Device complexity
If the device uses radar sensor data only for collision probability assessment, then the system remains simple, but the ability to identify unnecessary objects decreases leading to more false controls
Solution Approach 1:
The invention changes the parameters analyzed from radar sensor data by introducing spatial distribution patterns and movement characteristic analysis. Instead of using only basic collision probability parameters, the system examines additional parameters such as object density in specific regions and velocity patterns, enabling accurate identification of virtual image objects while maintaining system simplicity by using the same radar sensor
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
The device accurately identifies unnecessary objects, reducing the probability of unnecessary pre-collision controls and enhancing the accuracy of collision probability assessments.
Implementation Method 1
a sensor (20L, 20R) for transmitting a wireless medium and receiving the wireless medium reflected by an object to thereby detect the object
Data Source
AI summary
When an object detected by a sensor (20L, 20R) is not an unnecessary object and the object's collision probability index satisfies the predetermined condition, a pre-collision control ECU 10 performs a pre-collision control to prepare for a collision between an own vehicle and this object. The pre-collision control ECU determines that the object is the unnecessary object, when an initially detected location which is a location of the object in relation to the own vehicle when the sensor initially detected the object is in an unnecessary object area NA, and a velocity index value which represents a degree of a velocity change of an initial detected object whose the initially detected location is in the unnecessary object area does not fall within a predetermined range.