Autonomous Navigation Control Using Virtual Lanes and Grid Points
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous parking technologies face challenges in navigating narrow, one-way streets in parking lots, requiring complex calculations and high-performance calculators to manage turns, especially in limited spaces like building parking lots.
Innovation Solution
An apparatus and method for controlling autonomous navigation using parking lot map data with virtual lanes and grid points, where the autonomous vehicle adjusts its path based on the weight of each grid point to efficiently navigate and avoid obstacles, including pedestrians and reverse vehicles, by receiving and processing parking lot map data from a server.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If complex rotation track calculation is performed to enable left turns in narrow one-way streets, then the vehicle can navigate turns, but the calculator complexity and computational requirements increase significantly
Solution Approach 1:
The patent segments the continuous navigation path into discrete grid points arranged in virtual lanes. Instead of calculating complex continuous rotation tracks, the system divides the driving road into multiple virtual lanes with grid points at regular intervals, allowing the vehicle to navigate turns by selecting sequences of grid points rather than computing complex geometric rotation paths.
Solution Approach 2:
The patent introduces a virtual dimension by creating virtual lanes overlaid on the physical driving road. This virtual lane structure provides an additional layer of abstraction that simplifies turn navigation - the vehicle follows virtual lane guidance rather than directly computing rotation tracks in the physical space, effectively solving the turning problem through dimensional transformation.
2Measurement precision
If high-performance calculators are provided in the vehicle to perform complex calculations, then the navigation accuracy improves, but the device complexity and cost increase
Solution Approach 1:
By segmenting the navigation problem into discrete virtual lanes and grid points, the patent transforms complex continuous path calculation into simpler discrete point selection. This segmentation reduces computational requirements while maintaining navigation accuracy, as the system only needs to determine which grid points to follow rather than compute complex continuous trajectories.
Solution Approach 2:
The patent replaces the need for expensive high-performance calculators with a simpler computational approach based on pre-defined virtual lane structures. The virtual lane map data can be pre-processed and stored, allowing the vehicle's calculator to perform simpler real-time decisions about which grid points to follow, rather than performing complex real-time rotation track calculations.
3Ease of operation
If the vehicle follows a single center line in narrow driving roads, then the navigation is simple, but the vehicle cannot make left turns when needed
Solution Approach 1:
The patent segments the single center line concept into multiple virtual lanes, each with its own center line and grid points. This segmentation allows the vehicle to maintain simple center-line-following behavior within each virtual lane while providing the flexibility to switch between virtual lanes to execute left turns, thus preserving navigation simplicity while adding turning versatility.
Solution Approach 2:
The patent adds a virtual lane dimension to the physical driving road. This virtual dimension provides multiple parallel paths (virtual lanes) that the vehicle can select from, enabling left turns by switching to an adjacent virtual lane rather than attempting complex in-situ rotation maneuvers.
Data Source
AI summary
An apparatus for controlling autonomous navigation includes: a map data reception unit configured to receive parking lot map data indicating that a plurality of virtual lanes are located in a driving road in a parking lot and a grid point is located in each virtual lane by a certain interval from a parking server; and a controller configured to control driving of an autonomous vehicle based on a weight of each grid point included in the parking lot map data.