Reference Trajectory Generation by Clustering Normal Vehicle Paths
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing reference trajectory generation methods may produce inappropriate reference trajectories due to the inclusion of unnatural vehicle trajectories, which can affect autonomous driving systems, as these trajectories are not representative of normal vehicle travel patterns.
Innovation Solution
A reference trajectory generating device that classifies vehicle trajectories by clustering and selects the most common class to generate a reference trajectory, excluding abnormal or prohibited trajectories, using road structure data to refine the clustering process and averaging individual trajectories within the selected class.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If all vehicle trajectories are used for generating reference trajectory, then more trajectory data is available for processing, but the reference trajectory becomes inappropriate due to inclusion of unnatural trajectories
Solution Approach 1:
The patent extracts and removes unnatural trajectories from the dataset before generating the reference trajectory. By identifying and excluding trajectories that deviate from normal driving patterns (such as those with excessive lateral acceleration or improbable path deviations), the system ensures that only representative natural trajectories are used, thereby maintaining high reliability while processing a substantial quantity of valid data.
Solution Approach 2:
The patent applies different quality standards to different segments of trajectories. Rather than uniformly accepting or rejecting entire trajectories, the system evaluates local characteristics (such as lateral acceleration, curvature, and speed variations) at specific segments and selectively filters out only the unnatural portions, preserving the majority of valid trajectory data while eliminating problematic sections.
2Reliability
If clustering is applied to classify trajectories, then natural trajectories can be identified, but the processing complexity increases
Solution Approach 1:
The patent segments the trajectory classification process into distinct stages: first filtering trajectories based on basic naturalness criteria (lateral acceleration thresholds, speed ranges), then applying clustering algorithms only to the pre-filtered subset. This segmentation reduces the computational burden of clustering while maintaining accurate identification of natural trajectories.
Solution Approach 2:
The patent performs preliminary filtering and preprocessing of trajectories before applying clustering algorithms. By pre-processing the data to remove obviously unnatural trajectories and normalize the remaining data, the system reduces the complexity and computational resources required for the subsequent clustering step, while still achieving reliable identification of natural driving patterns.
Data Source
AI summary
A reference trajectory generating device includes a processor configured to classify a plurality of trajectories of travel of at least one vehicle through a predetermined section of a road into a plurality of classes by clustering of the trajectories in the predetermined section, select a class including the most trajectories of the classes, and generate a reference trajectory serving as a reference in the predetermined section by averaging individual trajectories included in the class selected from the classes.