3D Space-Time Trajectory Planning for Autonomous Parking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous and automated parking systems face computational challenges due to large search spaces and complex environments, leading to inefficient motion planning that requires numerous iterations between path and speed planners.
Innovation Solution
The system employs a 3D space-time artificial potential field for trajectory planning, allowing for a single iteration or a small number of iterations to determine an optimal trajectory by searching in a unified 3D space rather than iterating between 2D search planes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If iterative path and speed planning is used in complex parking environments, then collision-free trajectory can be achieved, but computational time becomes too long
Solution Approach 1:
The patent combines path planning and speed planning into a unified motion planning framework that operates in a 3D space-time search space. Instead of iteratively alternating between 2D path planning and 1D speed planning, the system performs both functions simultaneously in a single integrated search process, reducing computational iterations while ensuring collision-free trajectories.
Solution Approach 2:
The patent transitions from traditional 2D spatial search to a 3D space-time search by adding the time dimension to the search space. This allows the system to plan trajectories that simultaneously optimize spatial path and temporal speed profile, reducing the need for multiple iterative passes between path and speed planners.
2Device complexity
If traditional 2D search planes are used for motion planning, then path and speed can be planned separately, but the search process requires numerous iterations
Solution Approach 1:
The patent merges the separate 2D path search and 1D speed search into a single unified 3D space-time search. The motion planning system evaluates both spatial and temporal constraints simultaneously in one search process, eliminating the need for multiple iterative passes between path and speed planning stages.
Solution Approach 2:
The system elevates the search from two separate 2D planes (path space and speed space) to a unified 3D space-time search space. This dimensional integration allows simultaneous optimization of position and velocity along the trajectory, dramatically reducing computational iterations while maintaining planning quality.
Data Source
AI summary
This document describes trajectory planning in a three-dimensional (3D) search space with a space-time artificial potential field. An example system includes a processor that obtains an initial pose, a goal pose, and an obstacle map for an environment. The processor uses a parking trajectory algorithm to determine a trajectory by searching in a space-time artificial potential field. The trajectory includes a series of 3D waypoints, including two-dimensional (2D) positional coordinates and time coordinates, to navigate from the initial pose towards the goal pose. Operation of the host vehicle is then controlled to maneuver along the trajectory using an assisted-driving or autonomous-driving system. In this way, path searching is performed in the entire 3D search space rather than iterating between two 2D search planes. This allows motion planning for autonomous parking, especially with static and dynamic objects present, to determine an optimal trajectory in a single iteration.