Vehicle Trajectory Planning With Safe Terminal Stop Constraints
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current autonomous driving systems face challenges in planning stable and safe trajectories for self-driving vehicles, particularly at higher speeds, as they often sacrifice stability for safety or vice versa, and struggle to ensure safety for occupants and surrounding environments.
Innovation Solution
A method for trajectory planning that includes obtaining a reference trajectory and determining a back-up stop trajectory within a finite time horizon, forming a terminal set of states based on predefined constraints, and generating a nominal trajectory using a constraint-controlled technique like model predictive control (MPC), ensuring the nominal trajectory includes a terminal state for safety without incurring costs, thus balancing stability and safety.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a back-up stop trajectory is included in the trajectory planning to ensure safety, then the vehicle can reach a safe state, but the vehicle may deviate from the reference trajectory and lose stability
Solution Approach 1:
The trajectory planning is segmented into a nominal trajectory (for stability) and a back-up stop trajectory (for safety). The terminal set separates terminal states that can reach the safe state within the remaining time horizon from those that cannot. This segmentation allows the system to maintain nominal trajectory tracking while ensuring a safe backup option exists.
Solution Approach 2:
The back-up stop trajectory is pre-calculated and prepared in advance as part of the terminal set formulation. By determining the terminal set at the beginning of the prediction horizon and identifying which terminal states can reach the safe state, the system has safety measures ready before they are needed, allowing immediate switching without compromising stability.
2Stability of the object's composition
If the vehicle prioritizes tracking the reference trajectory for stability, then the vehicle maintains smooth operation, but the vehicle may not reach a safe state in emergency situations
Solution Approach 1:
The terminal set formulation provides feedback information about which terminal states can reach the safe state within the remaining time horizon. This feedback is incorporated into the MPC optimization, allowing the controller to adjust the nominal trajectory to ensure that at least one terminal state in the set can reach the safe state, thus maintaining both tracking performance and safety guarantee.
3Productivity
If the vehicle increases speed to improve productivity, then the vehicle covers more distance, but the vehicle has less time to reach a safe state in emergencies
Solution Approach 1:
The terminal set is dynamically adjusted based on the current state and remaining time horizon. As the vehicle speed increases and the time horizon decreases, the terminal set automatically adapts to reflect the reduced time available to reach the safe state. The MPC controller uses this dynamic terminal set to ensure that even at higher speeds, the vehicle can reach a safe state within the available time by adjusting the nominal trajectory accordingly.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present disclosure relates to a method for trajectory planning for a vehicle. The method comprises obtaining a reference trajectory over a finite time horizon, where the reference trajectory comprises a speed reference over time for the finite time horizon. Further, the method comprises determining a back-up stop trajectory within the finite time horizon. The back-up stop trajectory has a starting state and terminating in a final state, where the final state is defined as a safe state. The method further comprises forming a terminal set of states within the finite time horizon based on at least one predefined constraint, wherein the terminal set of states comprises a terminal state that corresponds to the starting state for the back-up stop trajectory. Moreover, the method comprises generating a nominal trajectory for at least a portion of the finite time horizon based on a constraint controlled technique, where the nominal trajectory is dependent on the obtained reference trajectory and a terminal constraint. The terminal constraint defines that the nominal trajectory comprises the terminal state. The constraint controlled technique comprises a cost minimizing control strategy, and the back-up stop trajectory from the starting state to the final state is associated with zero cost.