Vehicle Path Planning With Pose Constraints for Smooth Convergence
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional path planning systems for autonomous vehicles often result in rough vehicle motion and failed convergence to the target path due to initial vehicle offset, particularly in scenarios like automated valet parking, leading to comfort and user experience issues.
Innovation Solution
A method for path planning that incorporates a constraint based on the current vehicle pose, using two lateral boundaries extending from the vehicle's longitudinal center axis, which allows for smoother and more stable path generation by optimizing the path within a drivable area, formulated as a quadratic programming problem to derive a globally optimal solution in real time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional path planning generates a target path from the center of drivable area, then the path is optimized for the drivable area, but the vehicle experiences rough motion and failed convergence when initial offset is large
Solution Approach 1:
The patent applies preliminary action by adjusting the target path in advance based on the vehicle's initial pose before path tracking begins. The system calculates the initial offset between the vehicle and the generated path, then modifies the target path to account for this offset, preventing rough motion and convergence failure before they occur.
Solution Approach 2:
The patent changes the path parameters (target path coordinates and orientation) based on the vehicle's initial pose parameters. By dynamically adjusting the target path parameters according to the calculated initial offset, the system ensures smooth convergence regardless of the vehicle's starting position within the drivable area.
2Speed
If a large steering angle is commanded immediately after automation activation, then the vehicle can quickly converge to the target path, but the vehicle motion becomes rough and uncomfortable
Solution Approach 1:
The system performs preliminary calculation of the initial offset and adjusts the target path before path tracking begins. This preliminary action ensures that the vehicle starts tracking from an optimized position, enabling smooth and quick convergence without requiring large steering angles that would cause uncomfortable motion.
3Device complexity
If the target path is optimized around the center of drivable area, then the path planning is simplified, but the vehicle fails to converge when initial vehicle offset is very large
Solution Approach 1:
The patent maintains the simplicity of center-based path generation but adds a preliminary adjustment step that calculates the initial offset and modifies the target path accordingly. This preliminary action ensures convergence reliability without significantly increasing path planning complexity.
Solution Approach 2:
The system changes the target path parameters based on the vehicle's initial pose, dynamically adjusting the path to ensure convergence. This parameter adaptation maintains simplicity while improving reliability by accounting for large initial offsets.
Data Source
Figure 1
Figure 2
Figure 3a~3d
AI summary
The present disclosure relates to a method (400, 500) for path planning for an autonomous or semi-autonomous vehicle (1). The method (400, 500) comprises obtaining a drivable area of a surrounding environment of the vehicle, and generating a path (2) within the drivable area for a time step t based on a predefined set of characteristics for the path and a predefined set of constraints. The predefined set of constraints comprise at least one constraint (7, 8) based on a current pose of the vehicle.