Parking Trajectory Cost Grid for Real-Time Vehicle Maneuvers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for determining parking trajectories in autonomous vehicles are computationally expensive and often result in suboptimal and delayed results, making them unsuitable for real-time computation.
Innovation Solution
A parking management system that uses a three-dimensional grid containing parking costs to efficiently determine a parking trajectory. The grid is pre-computed and includes multiple layers associated with relative heading angles, allowing the vehicle to quickly access and aggregate parking costs for each state along a trajectory.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional trajectory determination techniques are used, then comprehensive trajectory evaluation can be achieved, but computational cost increases and processing time delays occur
Solution Approach 1:
The patent pre-computes and stores parking cost values for different vehicle states (positions, headings, velocities) in a cost map before real-time trajectory determination. During real-time operation, the system only needs to retrieve and aggregate these pre-computed costs along candidate trajectories, avoiding expensive on-the-fly computations while maintaining comprehensive evaluation capability
Solution Approach 2:
The patent divides the continuous state space into discrete grid cells, with each cell storing pre-computed parking cost information. This segmentation transforms the complex continuous optimization problem into a discrete lookup and aggregation task, significantly reducing real-time computational burden while preserving trajectory evaluation accuracy
2Reliability
If traditional trajectory determination techniques are used, then comprehensive trajectory evaluation can be achieved, but computational resources are excessively consumed
Solution Approach 1:
The system performs computationally intensive parking cost calculations in advance and stores results in a cost map. During real-time trajectory determination, the system only performs lightweight retrieval and aggregation operations, dramatically reducing energy consumption while maintaining comprehensive trajectory evaluation
Solution Approach 2:
The patent creates a simplified cost map representation that captures essential parking difficulty information without requiring full re-computation of complex trajectory dynamics. This copied representation enables fast evaluation with minimal computational resources
3Productivity
If real-time computation is required, then processing speed can be improved, but trajectory evaluation may become suboptimal
Solution Approach 1:
By pre-computing parking costs for all possible vehicle states and storing them in a cost map, the system enables real-time trajectory determination through simple aggregation operations, achieving both real-time speed and optimal evaluation quality simultaneously
Solution Approach 2:
The patent transforms the trajectory optimization problem from a continuous dynamic programming problem into a discrete grid-based cost aggregation problem. This dimensional transformation enables efficient real-time computation while preserving the comprehensive evaluation of trajectory quality through the cost map structure
Data Source
AI summary
Techniques for determining a parking trajectory for a vehicle are discussed herein. A parking management component may determine or receive a three-dimensional grid (“grid”), discretized based at least in part on a heading offset, a lateral offset, and/or a longitudinal offset between a first and second pose. A cell of the grid may include a cost associated indicating a minimum difference to the second pose when moving from the first as may be limited based on kinematic and/or dynamic constraints. When driving, a vehicle may determine a relative state of the vehicle to a desired location, use the relative state to access a cost from the grid, and determine whether to follow a trajectory based on the cost.