Autonomous Vehicle Motion Planning via S-L Coordinate Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicle motion planners, such as RRT and its variants, face challenges in complex environments due to increased search time and dimensionality from accounting for vehicle dynamics, often getting stuck or generating suboptimal solutions due to random sampling.
Innovation Solution
Separating path planning and velocity planning, using reference lines in combination with sampling-based motion planners, and projecting data into a S-L coordinate system to determine paths efficiently, with a dynamic vehicle look-up table to determine possible motions based on initial states and control inputs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If sampling-based motion planners (RRT and variants) are used for path planning in autonomous vehicles, then the search space is covered rapidly, but the solutions do not guarantee optimality and change in each planning cycle
Solution Approach 1:
The patent applies preliminary action by using a reference line (centerline) to pre-establish a guiding path before the sampling-based planner operates. The reference line is generated in advance based on road geometry and constraints, providing a predetermined optimal path that the RRT algorithm then samples around. This preliminary structure ensures solution optimality is preserved while allowing rapid exploration of the search space through random sampling.
2Reliability
If vehicle dynamics are accounted for in motion planning, then accurate vehicle motion is achieved, but the dimensionality of the search problem increases and search time increases
Solution Approach 1:
The patent segments the motion planning problem into two independent parts: (1) path planning in the configuration space using RRT to find the geometric path, and (2) velocity planning along the path using a separate velocity planner that incorporates vehicle dynamics. This segmentation allows the path planning to focus on spatial exploration without the computational burden of dynamics, while the velocity planner handles dynamics separately along the predetermined path, thus reducing overall search time while maintaining motion accuracy.
3Productivity
If the exploration rate of RRT is increased to cover search space rapidly, then path finding speed is improved, but the planner gets stuck or slows down in complex environments (e.g., curvy roads)
Solution Approach 1:
The patent introduces a reference line as an intermediary element that mediates between the random sampling of RRT and the complex environment constraints. The reference line acts as a guiding mediator that pre-processes the complex road geometry into a simplified centerline representation. This intermediary structure allows the RRT explorer to operate rapidly without getting stuck in complex curvy environments, as the reference line provides a predetermined safe path through the complexity.
Data Source
AI summary
Disclosed herein are methods and systems for motion planning in an autonomous vehicle (AV) that separate path planning and velocity planning and may use reference lines in combination with motion planners to determine paths. The method may include reference lines to project planning data into a S-L coordinate system. A motion planner algorithm uses the reference lines and previous path planning history to generate a path in the S-L coordinate system. A velocity is determined for the path. An AV controller is updated with the path and the velocity. Motion planning computations use a dynamic vehicle look-up table to determine possible vehicle motions based on an initial state and control input.