Automated Vehicle Trajectory Control via Reference Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for driver assistance systems struggle to determine and execute appropriate driving maneuvers in real time, especially in complex traffic situations involving multiple lane changes or prolonged periods, as they often fail to select the optimal control strategy among various possibilities.
Innovation Solution
A method that decouples driving maneuver planning and control, using a reference trajectory and generating possible trajectories with a non-linear mathematical model, selecting the most similar trajectory based on a cost functional that considers spatial distance and orientation, with weighting functions to prioritize certain segments and time segments, allowing for faster decision-making and reduced computational resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If previously known methods are used to determine driving maneuvers for longer periods or complex situations, then comprehensive control possibilities can be analyzed, but real-time determination is no longer possible
Solution Approach 1:
The patent divides the determination process into two independent segments: first determining a reference trajectory for the entire maneuver period, then generating possible trajectories only for a current predefined period. This segmentation allows each segment to be computed quickly while maintaining comprehensive analysis capabilities.
Solution Approach 2:
The reference trajectory is determined in advance for the entire maneuver period, serving as a preliminary guide. This preliminary action enables subsequent possible trajectories to be generated more efficiently by comparing against the pre-established reference, rather than analyzing all possibilities from scratch.
2Measurement precision
If a non-linear mathematical replacement model is used to generate possible trajectories, then accurate prediction of vehicle behavior in high acceleration situations is achieved, but computing time and resources increase
Solution Approach 1:
The patent applies different levels of model complexity to different parts of the trajectory determination process. The reference trajectory uses a simpler model, while possible trajectories use a more accurate non-linear model. This local differentiation optimizes the balance between accuracy and computing time.
Solution Approach 2:
Instead of using the complex non-linear model for the entire trajectory determination, the patent applies it only partially to generate possible trajectories from the reference trajectory. This partial application reduces overall computing time while maintaining accuracy where it matters most.
3Loss of time
If a simple point mass model is used for reference trajectory generation, then computing time is reduced, but detailed behavior reproduction is less relevant
Solution Approach 1:
The patent segments the modeling complexity: a simple point mass model is used for the reference trajectory where detailed behavior is less critical, while a more detailed non-linear model is used for possible trajectories where accurate behavior prediction is essential.
Solution Approach 2:
The patent changes the model parameters and complexity level according to the specific task: simpler parameters for reference trajectory generation, more detailed parameters for possible trajectory generation. This adaptive parameter selection optimizes both speed and accuracy.
Data Source
AI summary
A method for the automated control of a motor vehicle (10) is presented. The method includes the following steps:Generating and/or receiving a reference trajectory (T*) for the motor vehicle (10);Generating at least two possible trajectories ({circumflex over (x)}i) for the motor vehicle (10);Comparing the at least two possible trajectories ({circumflex over (x)}i) with the reference trajectory (T*); andSelecting one of the at least two possible trajectories ({circumflex over (x)}i) based on the comparison of the at least two possible trajectories ({circumflex over (x)}i) with the reference trajectory (T*).A control device (24) for a system for the control of a motor vehicle (10) is also proposed.