Grid Path Straightening for Obstacle-Aware Vehicle Routing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing path planning algorithms for autonomous vehicles face challenges in efficiently navigating around obstacles, particularly in large areas with multiple no-fly zones, due to high computational requirements and memory constraints, leading to latency and overhead issues.
Innovation Solution
The method involves determining an initial path using a grid-based system, converting between grid cells and real-world coordinates, and iteratively removing turn points if a straight line between them bypasses obstacles, resulting in a straightened path, with the use of local grids and shared grids to optimize processing and memory usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a grid-based path planning algorithm is used to navigate around obstacles in large areas, then the vehicle can avoid obstacles and follow a safe path, but the computational complexity and memory requirements increase significantly
Solution Approach 1:
The patent divides the large controlled area into multiple smaller local grids, each covering a specific sector. Instead of processing the entire large area as one grid, the system independently processes each local grid to generate path segments. This segmentation reduces the computational complexity and memory requirements for each processing step while maintaining complete obstacle avoidance coverage across the entire area.
2Reliability
If a comprehensive grid map of the entire area is created to ensure complete obstacle coverage, then all obstacles are accounted for, but the memory capacity required increases considerably
Solution Approach 1:
The patent creates multiple local grids that cover different sectors of the controlled area instead of one comprehensive grid. Each local grid contains obstacle information only for its specific sector, significantly reducing the memory capacity needed per grid. The system processes these smaller grids independently and combines their results to achieve complete obstacle detection across the entire area.
3Productivity
If traditional path planning algorithms process the entire path in one operation, then the complete path is generated, but latency and overhead issues occur
Solution Approach 1:
The patent generates the complete path by processing multiple local grids independently and sequentially. Each local grid processing operation completes quickly due to the smaller data size, and the individual path segments are then combined to form the complete path. This segmented approach reduces processing latency and overhead compared to processing the entire large-area grid in one operation.
Data Source
Figure 1
Figure 2
Figure 3A~3D
AI summary
Apparatuses and methods are disclosed for path planning. An apparatus comprising a data processor can be configured to determine an initial path past an obstacle between a first point and a second point, the initial path comprising at least two turn points between the first point and the second point. It can then be determined whether a line between two of the points and bypassing at least one of the turn points intersects the obstacle. A straightened path of travel can be determined in response to determination that the line between two of the points and bypassing the at least one of the turn points does not intersect the obstacle by removing the at least one bypassed turn point from the initial path.