Autonomous Vehicle Positioning via Road Boundary Deviation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in accurately positioning themselves on navigation maps, especially in new or unmarked road areas, as existing methods relying on landmarks are not effective.
Innovation Solution
A method and system that utilize a navigation map, approximate vehicle position and orientation, and environmental field of view to determine road boundaries and minimize angular and lateral deviations, allowing the vehicle to be precisely positioned on the map.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If landmark-based positioning is used, then positioning accuracy is improved in known territories, but the method fails completely in new unmarked road areas
Solution Approach 1:
The patent introduces road boundaries as an intermediary element that bridges the gap between known and unknown territories. Instead of relying on pre-stored landmarks that only exist in known areas, the system uses road boundaries detected by the sensor system as a mediator to establish positioning. The road boundary detection module identifies boundaries in the sensor field of view, and the road boundary matcher correlates these detected boundaries with map data to determine vehicle position, enabling operation in both familiar and new environments.
2Ease of operation
If approximate position is used for initial positioning, then the positioning process can start, but angular and lateral deviations reduce positioning accuracy
Solution Approach 1:
The patent implements a feedback mechanism where the sensor system continuously detects road boundaries in the vehicle's field of view, and the road boundary matcher compares these detections with expected boundaries from the navigation map. The deviation calculator quantifies the angular and lateral deviations between detected and expected boundaries, and this feedback is used to refine the vehicle's position estimate on the navigation map, progressively improving accuracy from the initial approximate position.
Solution Approach 2:
The patent replaces traditional mechanical positioning methods (which rely on pre-programmed landmarks and manual positioning) with an optical/sensor-based system. The sensor system captures visual information of the environment, the road boundary detection module processes this visual data to identify boundaries, and the matcher correlates these visual detections with map data, substituting mechanical positioning with an optical field-of-view-based approach that provides both ease of operation and improved accuracy.
Data Source
AI summary
Disclosed herein are a method and system for positioning an autonomous vehicle on a navigation map. The method includes positioning the autonomous vehicle on the navigation map, including receiving the navigation map, an approximate position and an approximate orientation of the autonomous vehicle on the navigation map, and an environmental field of view (FOV) of the autonomous vehicle and determining a first road boundary based on the navigation map and the approximate position of the autonomous vehicle, and a second road boundary based on the environmental FOV and the approximate orientation of the autonomous vehicle, further determining at least one of an angular deviation and a lateral deviation between the first road boundary and the second road boundary, and positioning the autonomous vehicle on the navigation map by minimizing at least one the angular deviation and the lateral deviation.