Vehicle Self-Position Estimation for Lane Count Change Sections
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing self-position estimation devices face inaccuracies in estimating the vehicle's position on a map when the number of lanes changes, as map link data is not accurately prepared for such sections, leading to potential erroneous estimations and identification of lane numbers.
Innovation Solution
A self-position estimation device equipped with a vehicle-mounted camera, vehicle state quantity sensor, satellite positioning unit, and map data storage, which recognizes changes in lane quantity and adjusts the weighting of the estimation position based on map data to stabilize accuracy, particularly in sections with unclear lane markings.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If map data is used for position estimation, then position estimation can be performed using pre-prepared link data, but estimation accuracy deteriorates in sections where lane quantity changes due to inadequate map link data preparation
Solution Approach 1:
The patent dynamically adjusts the weighting coefficient applied to map data-based position estimates based on the detected lane quantity. When lane quantity changes are detected, the weighting coefficient is reduced, allowing the system to adaptively balance between using convenient map data and maintaining accuracy in complex road sections
Solution Approach 2:
The system changes the parameter (weighting coefficient) of map data utilization based on road conditions. By modifying how map data is weighted rather than discarding it entirely, the system maintains ease of operation while improving accuracy in sections with lane quantity changes
2Measurement precision
If weighting of map data-based estimation is reduced in sections with lane quantity changes, then estimation accuracy is maintained, but system complexity increases due to additional recognition and weighting adjustment mechanisms
Solution Approach 1:
The patent segments the road into different sections based on lane quantity characteristics. By dividing the road into ordinary sections and sections with lane quantity changes, the system applies different weighting strategies to each segment, improving accuracy without requiring complete system redesign
Solution Approach 2:
The system introduces an intermediary mechanism (lane quantity recognition based on captured images) that mediates between map data and position estimation. This intermediary allows the system to maintain simple map data utilization while improving accuracy through conditional weighting adjustments
Data Source
AI summary
A self-position estimation device equipped to a vehicle: captures an image of a periphery of the vehicle; detects a state quantity of the vehicle; acquires position information indicating a position of the vehicle from a satellite system; stores map data that defines a map in which a road is expressed by a link and a node; estimate a self-position of the vehicle on the map, as an estimation position, based on the captured image, the state quantity, the position information, and the map data, respectively; recognizes a road section in which a lane quantity increases or decreases based on the captured image; and sets a weighting of the estimation position of the vehicle estimated based on the map data relatively smaller in response to a recognition of the road section in which the lane quantity increases or decreases.