Automated Parking Path Planning via Geometric-Kinematic Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing path planning methods for autonomous vehicle parking are limited to specific scenarios and geometries, failing to consider general parking scenarios and often require restrictive assumptions about obstacles and initial vehicle positions, making them impractical for real-time applications in varied parking environments.
Innovation Solution
A two-stage approach using a kinematic model to construct a collision-free path, followed by determining an actual trajectory, with the path planning problem partitioned into sub-problems and guided by a geometric path to increase computational efficiency and reduce unnecessary computations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If sampling-based path planning methods (such as RRT) are applied to automated parking, then the method can handle arbitrarily laid out parking spaces, but the computational cost becomes impractical due to unnecessary sampling of entire state space
Solution Approach 1:
The patent divides the path planning problem into two distinct stages: geometric path planning (ignoring kinematic constraints) and kinematic path planning (enforcing vehicle constraints). This segmentation allows the computationally intensive sampling to be performed only in the simplified geometric space rather than the full kinematic state space, dramatically reducing computational cost while maintaining versatility.
Solution Approach 2:
The patent introduces a geometric path as an intermediary representation that bridges the initial and target states without immediately enforcing complex kinematic constraints. This intermediate geometric path serves as a guide that is later refined into a kinematically feasible path, avoiding the need to sample the entire constrained state space directly.
2Productivity
If existing path planning methods are used for specific parking scenarios, then real-time path generation is achieved, but the methods are restricted to specific geometries and obstacle configurations
Solution Approach 1:
The patent creates a universal path planning framework that can handle multiple parking scenarios (parallel parking, perpendicular parking, angled parking, and arbitrary layouts) through a single two-stage approach. The geometric path planning stage is scenario-agnostic, and the kinematic refinement stage adapts to different vehicle constraints, making the system universally applicable without requiring scenario-specific algorithms.
3Reliability
If the entire parking space is sampled to find a collision-free path, then a feasible path can be found in complex parking spaces, but the computational time becomes impractical for real-time applications
Solution Approach 1:
The patent segments the path planning into geometric path finding (fast, ignores kinematics) and kinematic path refinement (slower, enforces constraints). By performing collision checking in the simplified geometric space rather than full kinematic space, the system achieves reliable collision-free paths with significantly reduced computational time.
Solution Approach 2:
The patent performs preliminary geometric path planning before kinematic refinement. This preliminary action establishes a collision-free geometric trajectory that guides subsequent kinematic sampling, preventing wasted computation on paths that would obviously violate constraints or collide with obstacles.
Data Source
Figure 1A
Figure 1B
Figure 1C
AI summary
The present invention relates to a method and a system for an automated parking determines, using the geometry of the vehicle and the map of the parking space, a collision free geometric path connecting an initial state of the vehicle with a target state of parked vehicle through a set of waypoints and determines, using a kinematic model of the vehicle, a set of kinematic subgraphs forming a kinematic graph having multiple nodes connected with kinematic edges. Each waypoint defines a position and orientation of the vehicle, each kinematic subgraph connects a pair of neighboring waypoints of the geometric path, each node defines a state of the vehicle, and each kinematic edge connecting two nodes defines a collision free kinematic path connecting the two nodes according to kinematics of the vehicle. A kinematic path is selected form the kinematic graph and a reference trajectory tracking the kinematic path as a function of time is determined using a dynamic model of the vehicle. The motion of the vehicle is according to the reference trajectory.