Automated Valet Parking Location Fail Detection with Adaptive Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Automated valet parking systems face challenges in accurately determining the location of autonomous vehicles due to errors from in-vehicle sensors, which can lead to navigation issues and potential collisions with parking boundaries.
Innovation Solution
An automated valet parking system that utilizes both in-vehicle sensors and facility sensors to acquire vehicle locations, with a location fail determiner setting different thresholds for longitudinal and lateral location errors based on passage width, node intervals, and node types to enhance detection sensitivity and prevent collisions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single threshold is used for both longitudinal and lateral location errors, then the system is simple to operate, but the measurement precision of location fail detection is insufficient
Solution Approach 1:
The patent applies local quality by setting different threshold values for different spatial dimensions (longitudinal and lateral directions). The system determines location fail based on whether the lateral error exceeds a first threshold value or the longitudinal error exceeds a second threshold value, where these threshold values differ based on the specific passage characteristics. This allows optimized detection precision for each direction without requiring a complex multi-factor decision system.
2Measurement precision
If a small lateral threshold is used to improve detection sensitivity, then location fail detection precision improves, but the system becomes more sensitive to normal variations causing false positives
Solution Approach 1:
The patent applies dynamics by making the threshold values adaptive rather than fixed. The system determines the first and second threshold values based on real-time factors including passage width, node interval distance, and curvature radius. When the passage width is smaller or the curvature radius is smaller, the threshold values are adjusted accordingly. This dynamic adjustment allows the system to maintain high detection sensitivity while accommodating normal variations in different passage conditions, thereby reducing false positives.
3Measurement precision
If location fail determination uses both longitudinal and lateral thresholds, then detection accuracy improves, but the computational complexity increases
Solution Approach 1:
The patent applies segmentation by dividing the location error evaluation into two independent dimensional assessments: lateral direction error comparison with the first threshold, and longitudinal direction error comparison with the second threshold. The system evaluates each dimension separately and determines location fail if either dimension exceeds its respective threshold. This segmented approach simplifies the computational logic compared to a combined multi-dimensional analysis, while maintaining comprehensive detection accuracy.
Data Source
AI summary
An automated valet parking system acquires a first vehicle location based on a detection result of an in-vehicle sensor of an autonomous vehicle and object information in a parking place, acquires a second vehicle location based on a detection result of a facility sensor provided in the parking place, and determines, based on the first and the second vehicle locations, presence or absence of location fail of the autonomous vehicle with respect to the first vehicle location. The location fail determiner determines that the location fail is present at least in one case out of a case where difference between a first longitudinal location and a second longitudinal location is equal to or greater than a longitudinal threshold or a case where difference between the first and the second lateral locations is equal to or greater than a lateral threshold. The lateral threshold is smaller than the longitudinal threshold.