Autonomous Vehicle Path Planning Without Precise Map Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Generating a travel path for autonomous vehicles is challenging in the absence of precise map information, as it requires accurate positioning and lane link data.
Innovation Solution
A method for autonomous driving that generates travel paths using state information to create preliminary and fixed sample points, determining optimal paths based on free space and heading conditions, and controlling the vehicle accordingly, without relying on precise map data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If precise map information is used for generating travel paths, then path generation accuracy is improved, but system complexity and data requirements increase
Solution Approach 1:
The patent extracts and removes the dependency on precise map information from the path generation system. Instead of requiring external map data, the system generates paths using only sensor data from the environment, effectively taking out the map information requirement and reducing system complexity while maintaining path generation capability
Solution Approach 2:
The system performs self-service by generating its own path planning data without external map inputs. The vehicle uses its sensors to perceive the environment and autonomously generates travel paths based on detected obstacles, road boundaries, and spatial relationships, making the system self-sufficient and independent of complex map data
2Device complexity
If traditional path generation methods are used without map information, then system simplicity is improved, but path generation capability deteriorates
Solution Approach 1:
The system performs preliminary actions by pre-processing sensor data to identify navigable spaces, obstacles, and path candidates before actual path generation. This preliminary processing of environmental data enables reliable path generation without maps, as the system has already organized spatial information and identified viable routing options in advance
Solution Approach 2:
The patent transitions from two-dimensional map-based path planning to three-dimensional spatial reasoning using sensor data. By utilizing depth information, vertical clearance data, and multi-sensor fusion, the system creates paths in 3D space without requiring 2D map representations, thereby maintaining simplicity while improving capability
3Manufacturing precision
If multiple sample points are generated and evaluated, then path optimization is improved, but computational time increases
Solution Approach 1:
The patent segments the path generation process into distinct stages: generating multiple preliminary sample points, evaluating them against constraints, selecting optimal points, and refining the path. This segmentation allows the system to efficiently manage computational resources by processing points in batches and eliminating poor candidates early, reducing overall computational time while maintaining optimization quality
Solution Approach 2:
The system generates a limited number of sample points that is sufficient for finding good paths without exhaustively searching all possible paths. By using heuristic methods to generate only promising candidate points and evaluating them partially through constraint checking, the system achieves adequate path optimization without the excessive computational time required for complete enumeration
Data Source
AI summary
A method performed for autonomous driving of a vehicle is introduced. The method may comprise generating, based on state information of a moving object, a point in a first sampling area as a first preliminary sample point, determining, based on a sample point fixing condition associated with a goal point and information associated with the first preliminary sample point, a first fixed sample point, generating, based on updated state information, a point in a second sampling area as a second preliminary sample point, determining, based on the sample point fixing condition and information associated with the second preliminary sample point, a second fixed sample point, generating, based on a fixed sample point group comprising the first fixed sample point and the second fixed sample point, a travel path, generating, based on the travel path, a control signal, and controlling, based on the control signal, the vehicle for autonomous driving.