Autonomous Vehicle Trajectory Selection for Adaptive Maneuvering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for generating trajectories for autonomous vehicles are inefficient and costly, as they often involve generating a large number of random trajectories without considering current circumstances, leading to suboptimal maneuvering behaviors.
Innovation Solution
A method where processors select and filter maneuvering options based on predetermined behaviors, generate trajectories that satisfy specific constraints, and choose the trajectory with the lowest cost to avoid collisions and minimize uncomfortable maneuvers, allowing the vehicle to adapt its behavior dynamically.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a large number of random trajectories are generated in a predetermined period of time, then the coverage of possible maneuvering behaviors is improved, but the computational efficiency and cost increase significantly
Solution Approach 1:
The system preselects a limited set of relevant maneuvering options based on the current driving situation before trajectory generation. This preliminary filtering of maneuvering options (e.g., lane changes, turns, straight continuation) reduces the search space from potentially infinite random trajectories to a manageable set of contextually relevant options, thereby improving computational efficiency while maintaining adequate coverage of necessary maneuvering behaviors
Solution Approach 2:
The system generates trajectories with different levels of detail and complexity based on local driving conditions. For simple situations, fewer trajectories are generated with basic maneuvering options, while complex situations trigger generation of more trajectories with additional maneuvering variations. This localized adaptation of trajectory generation intensity optimizes computational resources according to actual needs
2Productivity
If predetermined maneuvering options are preselected for trajectory generation, then the computational cost is reduced, but the adaptability to dynamic situations decreases
Solution Approach 1:
The system dynamically adjusts the set of maneuvering options based on real-time driving conditions, objects detected by the perception system, and predicted behaviors of other road users. The maneuvering options are not fixed but are selected and filtered adaptively according to the current situation, allowing the system to maintain high computational efficiency while responding appropriately to dynamic changes in the environment
Solution Approach 2:
The system uses feedback from the perception system and cost analysis to iteratively refine trajectory selection. After generating trajectories from preselected maneuvering options, the system evaluates their costs and can trigger regeneration of trajectories with different maneuvering options if the initial selections prove suboptimal, thereby maintaining adaptability through iterative refinement
3Reliability
If multiple trajectories are generated and analyzed for cost, then the quality of maneuvering behavior is improved, but the time required for trajectory generation increases
Solution Approach 1:
The system changes parameters such as the number of trajectories generated, the complexity of maneuvering options, and the detail of cost analysis based on driving conditions. In low-risk situations, fewer trajectories with simpler evaluation are sufficient, reducing generation time. In high-risk or complex situations, the system increases trajectory count and evaluation depth, accepting longer generation times when necessary for safety
Data Source
AI summary
Aspects of the disclosure provide for generation of trajectories for a vehicle driving in an autonomous driving mode. In one instance, a default number of trajectories to be generated may be identified. A set of maneuvering options may be selected from a set of predetermined maneuvering options based on the number of trajectories. The set of maneuvering options may be filtered based on the default number of trajectories. A set of trajectories may be generated based on the filtered set of maneuvering option such that each trajectory of the set corresponds to a different maneuvering behavior. A cost for each trajectory of the set of trajectories may be determined, and one of the trajectories of the set of trajectories may be selected based on the determined costs. The vehicle may be maneuvered in the autonomous driving mode according to the selected one of the trajectories.