Vehicle Trajectory Determination via Grid Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining road vehicle trajectories are computationally intensive, making them inefficient for real-time driving assistance and collision risk assessment.
Innovation Solution
A method that determines at least two reference trajectories for a vehicle, calculates a reference kinematic profile for each, and uses these to interpolate a vehicle time slice, allowing for the determination of information about an intermediate trajectory without explicit calculation of the intermediate trajectory itself, leveraging homotopy properties to reduce computational load.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a large number of possible trajectories are generated and compared to determine desired information, then the information is useful for driving the vehicle, but the computation time becomes excessively long
Solution Approach 1:
The patent segments the continuous trajectory space into a discrete grid representation. Instead of evaluating all possible continuous trajectories, the method divides the space into grid cells and represents trajectories as sequences of grid cells. This segmentation reduces the infinite trajectory space to a manageable discrete set, enabling efficient computation while preserving essential trajectory information for driving decisions.
Solution Approach 2:
The patent pre-calculates and stores the grid representation of the environment and pre-determines valid grid-based trajectories before actual driving decisions are needed. By performing this computational work in advance and storing the results, the system avoids repeating expensive trajectory calculations during time-critical driving situations, thus reducing real-time computation time while maintaining information quality.
2Measurement precision
If all possible trajectories are explicitly calculated to assess collision risk, then the collision risk assessment is accurate, but the computational complexity increases significantly
Solution Approach 1:
The patent creates a simplified copy of the continuous trajectory space in the form of a discrete grid model. Instead of working with complex continuous trajectory equations, the method uses grid cell sequences as simplified representations that capture the essential spatial relationships and collision risks. This copying approach maintains assessment accuracy while dramatically reducing computational complexity.
Solution Approach 2:
The patent transforms the problem from continuous trajectory space to a discrete grid-based representation, effectively changing the dimensional nature of the computation. By mapping continuous coordinates to discrete grid cells, the method reduces the complexity of trajectory comparison and collision detection operations, making them computationally tractable while preserving the essential geometric relationships needed for accurate risk assessment.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The system has a device for identifying a moving object e.g. pedestrian, relative to a road vehicle (12), and estimating a position of the moving object. A calculation unit determines beams (F12, F25, F26) of trajectories (T1, T2) for the vehicle that is likely to traverse for a future period, kinematic profiles for the vehicle during a displacement of the vehicle on one of the trajectories, and a probability of the kinematic profiles. The calculating unit determines a probability of collision between the moving object and the vehicle based on the probability of the kinematic profiles.