Vehicle Sensor Fusion Track Alignment for Collision Point Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current autonomous driving sensor fusion systems inaccurately represent the closest point of potential collision between vehicles due to mismatches between LiDAR and other sensing devices' data, leading to delayed longitudinal control and potential collisions.
Innovation Solution
A method and system that determine the closest point of potential collision by aligning LiDAR and sensor fusion tracks through identifying corner positions based on heading angles, midpoints, and coordinate values, adjusting the sensor fusion track to improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sensor fusion track is generated by selecting and processing data from multiple sensing devices, then comprehensive target object information is obtained, but the position accuracy of the closest point of potential collision deteriorates due to mismatches between LiDAR and other sensing devices' data
Solution Approach 1:
The patent uses LiDAR track data as an intermediary reference to correct the sensor fusion track. By comparing the closest point positions between LiDAR track and sensor fusion track, the system identifies mismatches and applies corrections to align them, thereby improving position accuracy without fundamentally changing the multi-sensor fusion architecture
Solution Approach 2:
The system implements a feedback mechanism where the LiDAR track serves as a reference standard to evaluate and correct the sensor fusion track. The closest point position from LiDAR is used as feedback to adjust the sensor fusion results, creating a closed-loop correction process that continuously improves accuracy
2Reliability
If the closest point position on sensor fusion track differs from actual target vehicle position, then collision risk assessment is delayed, but adjusting the track requires additional processing complexity
Solution Approach 1:
The patent extracts the critical correction information by identifying only the closest point position mismatch between LiDAR and sensor fusion tracks. Rather than reprocessing entire track data, the system focuses on extracting and correcting the specific positional discrepancy, thereby reducing processing complexity while improving reliability
Solution Approach 2:
The system performs preliminary alignment by using LiDAR track data to pre-correct the sensor fusion track before collision risk assessment. By adjusting the closest point position in advance based on LiDAR reference data, the system ensures accurate collision risk evaluation without requiring complex real-time adjustments
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
Enhances the accuracy of the closest point of potential collision on the sensor fusion track, reducing the risk of collisions by aligning it with the LiDAR track's prediction, thereby improving autonomous driving control.
Implementation Method 1
determining a first point corresponding to a closest point of a target object from the vehicle with respect to a potential collision based on a Light Detection and Ranging (LiDAR) track thereof
Data Source
AI summary
A method for sensor fusion for a vehicle, includes determining a first point corresponding to a closest point of a target object from the vehicle with respect to a potential collision based on a LiDAR track thereof and a heading of the vehicle, determining a second point corresponding to a closest point of the target object from the vehicle with respect to a potential collision based on a sensor fusion track, and updating the sensor fusion track based on the first point and the second point.