Autonomous Vehicle Passable Area Planning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous driving vehicles face challenges in determining a safe passable area when encountering obstacles, often resulting in the frontend of the vehicle being too close to obstacles due to traditional discrete sample point planning methods.
Innovation Solution
The vehicle determines the lane configuration, obstacle boundary, and passable area based on the size and shape of the obstacle, allowing expansion into adjacent lanes if necessary, and uses predicted trajectories for dynamic obstacles, with optimization techniques like quadratic programming to plan a safe trajectory.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional discrete sample point planning methods are used, then the planning process is simple, but the frontend of the vehicle becomes too close to obstacles
Solution Approach 1:
The planning space is segmented into discrete sample points arranged in a grid pattern, where each point represents a possible vehicle state (position and orientation). This segmentation allows the complex continuous planning problem to be broken down into manageable discrete states that can be evaluated individually for safety and feasibility.
Solution Approach 2:
The planning approach transitions from traditional 2D spatial planning to a 4D state space planning by adding orientation and time dimensions. The sample points are distributed not only in spatial coordinates (x, y) but also include vehicle orientation (θ) and time (t), creating a comprehensive state space that ensures safer vehicle-obstacle separation.
2Adaptability or versatility
If the vehicle expands passable area into adjacent lanes, then the ability to pass obstacles is improved, but the risk of exceeding lane boundaries increases
Solution Approach 1:
The passable area boundaries are dynamically adjusted based on the vehicle's predicted trajectory, speed, and orientation at each sample point. Rather than using fixed lane boundaries, the system adapts the effective boundaries by incorporating vehicle dimensions and motion characteristics, allowing the vehicle to safely utilize adjacent lanes when necessary while maintaining compliance through continuous boundary reevaluation.
Solution Approach 2:
The system performs preliminary evaluation of sample points near lane boundaries by checking whether the vehicle would exceed boundaries at future time steps. By proactively identifying and excluding boundary-exceeding sample points before trajectory generation, the system ensures lane compliance is maintained while still allowing flexible use of adjacent lanes when safe.
3Reliability
If the vehicle accounts for its size and shape in trajectory planning, then collision avoidance is improved, but the computational complexity increases
Solution Approach 1:
The vehicle body is segmented into discrete grid cells matching the sample point resolution, and collision detection is performed by checking whether the vehicle's occupied cells (based on its dimensions and orientation) intersect with obstacle cells. This segmentation approach efficiently handles vehicle size and shape considerations without requiring complex continuous geometric calculations at each sample point.
Solution Approach 2:
Instead of directly calculating the vehicle's complex geometric shape and its interaction with obstacles, the system creates a simplified discrete representation (copy) of the vehicle's occupied space on the grid. This copy, determined by the vehicle's length, width, and orientation at each sample point, enables efficient collision detection while capturing the essential spatial constraints.
Data Source
AI summary
According to one embodiment, in response to determining that an obstacle blocks at least a portion of a current lane in which an ADV is driving, an obstacle boundary of the obstacle is determined based on the size and shape of the obstacle. A lane configuration is determined based on map data of a map corresponding to a road associated with the lanes. A passing lane boundary that can be utilized by the ADV is determined based on the lane configuration of the road and the obstacle boundary of the obstacle. A passable area is calculated within the passing lane boundary based on a size of the ADV. The passable area is utilized by the ADV to pass the obstacle without collision. Thereafter, a trajectory is planned within the passable area boundary to control the ADV to pass the obstacle.