Autonomous Driving Trajectory Correction for Complex Road Scenes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing trajectory generation technologies struggle to handle complex driving scenes, such as unexpected obstacles like cattle or sheep on the road, leading to poor correction effects and reduced accuracy in autonomous driving.
Innovation Solution
A method involving a target trajectory generation model and a target driving correction model that utilize driving-related data to generate and correct candidate driving trajectories, incorporating sensor information, navigation planning, and driving rules, with a visual language model for scene analysis and correction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If an end-to-end model is used for trajectory correction, then the system is simple to implement, but it cannot handle complex driving scenes effectively
Solution Approach 1:
The system divides the trajectory generation task into two independent modules: a trajectory generation model for normal scenes and a trajectory correction model for complex scenes. This segmentation allows each module to specialize in its function, with the correction model specifically trained to handle complex driving scenes while the generation model handles routine operations, thereby resolving the contradiction between implementation simplicity and complex scene handling capability.
2Ease of operation
If manual correction is used for driving trajectory, then the system is easy to understand, but the correction effect is poor and accuracy is reduced
Solution Approach 1:
The trajectory correction model implements a feedback mechanism where the generated trajectory from the generation model is input into the correction model, which then outputs corrected trajectory information. This feedback loop continuously refines the trajectory by comparing predicted paths with actual driving conditions, significantly improving correction accuracy while maintaining system simplicity through automated processing.
3Measurement precision
If a dual-model system is implemented for trajectory generation and correction, then the accuracy for complex scenes is improved, but the device complexity increases
Solution Approach 1:
The trajectory generation model and correction model are merged into a unified processing framework where the generation model outputs candidate trajectories that are then refined by the correction model. This merging approach allows the system to leverage the strengths of both models while sharing common infrastructure such as input data processing and output integration, thereby improving complex scene handling accuracy without proportionally increasing system complexity.
Data Source
AI summary
The present disclosure discloses a trajectory generation method, apparatus, electronic device, storage medium and program, wherein the method includes: acquiring driving-related data corresponding to a current vehicle; inputting first driving-related data in the driving-related data to a pre-created target trajectory generation model to obtain a candidate driving trajectory corresponding to the current vehicle; inputting second driving-related data in the driving-related data into a pre-created target driving correction model to obtain corresponding driving correction information; and correcting the candidate driving trajectory based on the driving correction information to obtain a corresponding target driving trajectory, such that the current vehicle automatically drives according to the target driving trajectory. The present disclosure corrects the candidate driving trajectory by the driving correction information to obtain the target driving trajectory, effectively avoiding the problem of poor correction effect caused by manual trajectory correction, and improving the accuracy and effectiveness of trajectory correction.