Autonomous Vehicle Path Centering With Random Lateral Shift
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous driving vehicles consistently using the same part of the road can lead to damage to the road surface, as their deterministic path planning lacks variability.
Innovation Solution
A random shift-based path centering system that generates a lateral shift distance value to shift the reference line of the vehicle's trajectory, allowing it to navigate with a randomized trajectory that avoids constant usage of the same road section.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If deterministic path planning is used for autonomous vehicles, then navigation accuracy is improved, but road surface damage increases due to repeated usage of the same path
Solution Approach 1:
The system dynamically adjusts the vehicle's trajectory by introducing random lateral shifts to the reference line. Instead of following a fixed deterministic path, the vehicle's trajectory becomes variable while still maintaining accurate navigation to the destination. The random shift module modifies the reference line dynamically based on generated shift values, resolving the contradiction between navigation accuracy and road surface protection.
Solution Approach 2:
The system changes the positional parameters of the vehicle's trajectory by applying random lateral shifts to the reference line coordinates. This parameter modification allows the vehicle to deviate from the exact center of the road while maintaining overall navigation accuracy, thereby distributing wear across different road sections and preventing concentrated damage at fixed path locations.
2Object-affected harmful factors
If random lateral shift is applied to the trajectory, then road surface wear is distributed evenly, but navigation precision may be reduced
Solution Approach 1:
The random shift is applied locally to the reference line at different segments of the trajectory rather than uniformly across the entire path. This allows the vehicle to distribute wear across multiple road sections while maintaining accurate navigation overall. Each local segment experiences a controlled deviation, preventing concentrated wear without compromising the global navigation precision to the destination.
Data Source
AI summary
Embodiments of the present disclosures disclose a method and a system to generate a path planning trajectory with a random lateral shift for an autonomous driving vehicle (ADV). In one embodiment, a system generates a reference line to navigate the ADV from a start location to a destination location. The system determines a lateral shift distance value to shift a lane center for the reference line. The system generates a shifted trajectory using the reference line based on the lateral shift distance value. The system controls the ADV based on the shifted trajectory to navigate the ADV.