Left-Turn Path Planning Using Rational Bezier Curves at Intersections
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The complexity of intersections in city streets poses challenges for safe and accurate path planning in autonomous vehicles, as they involve diverse vehicle behaviors, scenarios, and varying intersection sizes and types, making it difficult to ensure traveling safety.
Innovation Solution
An apparatus and method utilizing sensors like cameras and radars to generate a grid map, select control points based on minimum turning radii, and create rational Bezier curves to plan a safe reference path for vehicles, adjusting curvature and safety distances to navigate intersections effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If path planning logic is used for autonomous traveling in city streets, then traveling safety is improved, but the system cannot handle diverse intersection scenarios and vehicle behaviors effectively
Solution Approach 1:
The patent segments the path planning process into distinct modules: grid map generation from precision maps, intersection type classification, control point selection based on minimum turning radius, and rational Bezier curve generation. This modular segmentation allows each component to be optimized independently for both safety and adaptability to diverse intersection scenarios.
Solution Approach 2:
The system dynamically changes parameters such as control point positions, turning radius, and Bezier curve weights based on the specific intersection type and opposing vehicle conditions. This parameter adaptation enables the same path planning framework to handle diverse intersection scenarios while maintaining traveling safety through mathematically optimized paths.
2Ease of manufacture
If a fixed path planning method is used, then the system is simple to implement, but it cannot adapt to different intersection types and sizes
Solution Approach 1:
The patent creates a universal path planning system that handles multiple intersection types (cross intersections, T-intersections, Y-intersections, five-way intersections, six-way intersections) using a single unified approach. The system determines intersection type from precision map data and applies appropriate control point selection and Bezier curve generation rules for each type, achieving multi-functionality without requiring separate planning algorithms for each scenario.
Solution Approach 2:
The system dynamically adjusts path planning parameters based on real-time conditions including intersection geometry, vehicle dimensions, and opposing vehicle turning radius. The control points and Bezier curve weights are calculated dynamically rather than being fixed, allowing the same implementation framework to adapt to different intersection types and sizes while maintaining mathematical rigor and safety guarantees.
3Productivity
If the host vehicle follows the minimum turning radius path, then the turning maneuver is efficient, but it may conflict with opposing left-turn vehicles
Solution Approach 1:
The patent introduces an intermediary mathematical model (rational Bezier curve) that mediates between the host vehicle's turning efficiency and conflict avoidance requirements. The Bezier curve is constructed with control points positioned at calculated distances from the minimum turning radius path, creating a safe intermediate trajectory that maintains efficiency while preventing conflicts with opposing left-turn vehicles through controlled curvature adjustment.
Solution Approach 2:
The system performs preliminary calculation of control point positions and Bezier curve parameters before the actual turning maneuver. By pre-calculating the optimal path that balances turning efficiency with safety margins from opposing vehicles, the system eliminates the need for reactive adjustments during the turn, maintaining both productivity and reliability throughout the maneuver.
Data Source
AI summary
Disclosed herein is an apparatus for driver assistance including a sensor installed in a host vehicle, including at least one of a camera or a radar, and configured to output data; a memory in which a precision map including intersection information is stored, and at least one processor electrically connected to the memory and configured to process the data output from the sensor. The at least one processor may generate a grid map by connecting lane lines of an intersection on the precision map when the host vehicle makes a left turn at the intersection, select a start point and an end point required for the left turn on the grid map, select at least one control point according to a minimum turning radius of the host vehicle or an opposite left-turn vehicle, generate a rational Bezier curve based on the start point, the end point, the at least one control point, and determine the rational Bezier curve to be a left-turn path that the host vehicle follows, and control the host vehicle based on the processing of the data such that the host vehicle follows the left-turn path.