Autonomous Vehicle Lane Change Trajectory Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous driving vehicles face challenges in efficiently planning and executing lane changes without significant overshooting, lacking effective methods to navigate lane changes quickly and smoothly.
Innovation Solution
The system generates a first trajectory from the current lane to a target lane, calculates a lane shifting correction based on lane configurations and the vehicle's state, and modifies the starting point of the trajectory to generate a second trajectory, which is used to control the vehicle's lane change, with additional steering commands to compensate for drifting errors and adjust the lane change time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a standard trajectory is generated for lane changing, then the vehicle can follow a predictable path, but the trajectory may cause overshooting or drifting errors
Solution Approach 1:
The system continuously monitors the vehicle's actual position during lane changing and compares it with the planned trajectory. When drifting errors are detected, the system dynamically adjusts the trajectory in real-time to compensate for deviations, ensuring the vehicle reaches the target lane accurately without overshooting.
Solution Approach 2:
The trajectory is not fixed but dynamically adjusted based on the vehicle's current state, lane configuration, and detected errors. The system modifies trajectory parameters such as curvature and transition speed adaptively during the lane changing process to maintain precision while following a predictable overall path.
2Manufacturing precision
If the lane changing process is extended to reduce overshooting, then trajectory accuracy improves, but the time required for lane changing increases
Solution Approach 1:
The system pre-calculates the optimal trajectory considering lane width, curvature, and vehicle dynamics before initiating lane changing. By planning the entire path in advance with appropriate safety margins, the system achieves accurate lane changing in minimal time without requiring extended correction phases.
Solution Approach 2:
The system adjusts trajectory parameters such as lateral acceleration, steering angle, and transition duration based on real-time conditions. By optimizing these parameters dynamically, the system achieves both high precision and fast execution, adapting the lane changing speed and smoothness to match the specific situation.
3Manufacturing precision
If the trajectory is adjusted frequently to compensate for errors, then lane changing precision improves, but the control complexity increases
Solution Approach 1:
The system applies corrections selectively at critical phases of lane changing rather than continuously adjusting all parameters. Specifically, the trajectory is modified at the starting point and key transition points based on detected errors, rather than recalculating the entire trajectory frequently, thus maintaining precision while limiting control complexity.
4Adaptability or versatility
If the starting point of the trajectory is shifted to account for lane configuration, then the vehicle can adapt to different lane widths and curvatures, but the trajectory planning complexity increases
Solution Approach 1:
The system pre-adjusts the trajectory starting point based on detected lane configuration parameters such as lane width, curvature, and markings before initiating lane changing. By incorporating these adaptations in advance, the system handles diverse lane configurations effectively without requiring complex real-time adjustments during execution.
Data Source
AI summary
In one embodiment, during a planning stage of autonomous driving of an autonomous driving vehicle (ADV), it is determined that the ADV needs to change lanes from a source lane to a target lane. A first trajectory is generated from a current location of the ADV in the source lane to the target lane such as a center line of the target lane. A lane shifting correction is then calculated based on the lane configuration of at least the source lane and/or target lane, as well as the current state of the ADV. Based on the lane shifting correction, at least the starting point of the first trajectory is modified, which in turn generates a second trajectory. In one embodiment, the starting point of the first trajectory is shifted laterally with respect to a heading direction of the source lane based on the lane shifting correction.