Autonomous Vehicle Parking Path Planning with Dynamic Obstacle Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current parking logic in autonomous driving vehicles (ADVs) does not account for the movement of obstacles when reversing, leading to potential collisions and computationally expensive solutions that may not produce a parking path within an acceptable time frame.
Innovation Solution
A computer-implemented method for ADVs that determines parking space constraints and generates parking paths considering kinematic constraints, then adjusts these paths by analyzing and predicting the movement of obstacles using a perception and planning system, allowing for obstacle avoidance and efficient path optimization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If parking logic treats obstacles as static objects, then the parking algorithm is simpler to implement, but collisions with moving obstacles cannot be avoided
Solution Approach 1:
The patent transforms the static obstacle model into a dynamic one by incorporating obstacle speed and direction predictions. The system now considers moving obstacles as dynamic entities with predicted trajectories, allowing the parking algorithm to adjust paths in real-time to avoid collisions while maintaining reasonable computational complexity through efficient prediction models.
Solution Approach 2:
The system performs preliminary prediction of obstacle movement before path planning. By predicting where obstacles will be during the parking maneuver rather than treating them as static, the system proactively identifies potential collision risks and adjusts the parking path in advance, preventing collisions before they occur.
2Reliability
If computationally expensive solutions are used to account for obstacle movement, then collision avoidance improves, but parking time exceeds acceptable delays
Solution Approach 1:
The patent divides the parking problem into segmented phases: obstacle prediction, initial path planning, and path validation/adjustment. This segmentation allows the system to use computationally efficient methods for prediction and initial planning, then apply more rigorous collision checking only to the specific paths generated, reducing overall computation time while maintaining safety.
Solution Approach 2:
The system applies partial collision checking focused on critical path segments rather than exhaustive checking of entire paths. By identifying and prioritizing checking of high-risk segments where moving obstacles are most likely to intersect with the parking path, the system achieves adequate collision avoidance with reduced computational effort compared to complete path verification.
3Productivity
If parking paths are planned without considering obstacle speed and direction, then computational cost is reduced, but the parking path may conflict with moving obstacles
Solution Approach 1:
The system performs preliminary prediction of obstacle trajectories before path planning. By calculating where obstacles will be positioned during the parking maneuver in advance, the system incorporates this predictive information into path generation, ensuring paths are planned with awareness of moving obstacle locations without requiring complex real-time optimization during path execution.
Data Source
AI summary
A parking system for autonomous driving vehicles optimizes a solution to a parking problem. The ADV detects a parking lot and selects a parking space. The ADV defines constraints for the parking lot, parking space, and kinematic constraints of the ADV, and generates a plurality of potential parking paths to the parking space, taking into account the constraints of the parking lot, parking space, and kinematics of the ADV, but without taking into any obstacles that may be surrounding the ADV. The ADV determines a cost for traversing each of the parking paths. One or more least cost candidate paths are selected from the parking paths, then one or more candidate paths are eliminated based on obstacles surrounding the ADV. Remaining candidates can be analyzed using a quadratic optimization system. A best parking path can be selected from the remaining candidates to navigate the ADV to the parking space.