Vehicle Trajectory Planning Using Time-Varying Reference
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomously driving vehicles face challenges in efficiently and reliably tracking a desired path, particularly in lane centering applications, as existing methods often result in jerky or unstable steering responses due to constant reference values throughout the prediction horizon, leading to discomfort and potential collisions.
Innovation Solution
A computer-implemented method for vehicle trajectory planning that determines a varying reference trajectory based on time, using a tanh-function, initial lateral offset, arc-length state, and vehicle velocity, which is calculated using sensors and model predictive control to generate a smoother steering command.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a constant reference value is used throughout the prediction horizon, then the control implementation is simple, but the steering response becomes jerky and unstable
Solution Approach 1:
The patent applies dynamics by transitioning from a static constant reference value to a dynamic time-varying reference trajectory. The reference trajectory evolves over the prediction horizon based on the tanh-function, allowing the system to adapt smoothly to changing vehicle states while maintaining computational tractability through a structured functional form.
Solution Approach 2:
The patent changes the parameter of the reference value from constant to time-varying. Specifically, the reference trajectory is defined as a function of time and vehicle state (lateral offset, arc-length, velocity), transforming the control reference from a fixed parameter to a dynamic parameter that adapts to the vehicle's motion state, thereby smoothing the steering response.
2Stability of the object's composition
If a varying reference trajectory is used, then the steering response becomes smoother and more stable, but the computational complexity increases
Solution Approach 1:
The patent introduces dynamics into the reference trajectory to improve stability, but constrains the functional form to maintain computational efficiency. The tanh-function-based trajectory provides smooth transitions while requiring only basic arithmetic operations and function evaluations, avoiding the need for complex optimization or iterative calculations.
Solution Approach 2:
The patent uses parameter changes to define the reference trajectory in terms of vehicle state variables (lateral offset, arc-length, velocity). This approach allows the varying reference to adapt to real-time conditions while being computed from readily available sensor measurements, balancing trajectory accuracy with computational feasibility.
3Device complexity
If the reference trajectory does not account for vehicle velocity, then the control logic is simpler, but the trajectory accuracy and passenger comfort deteriorate
Solution Approach 1:
The patent incorporates vehicle velocity as a parameter in the reference trajectory definition. The velocity term scales the reference trajectory appropriately for different driving conditions, ensuring accurate lane centering behavior whether the vehicle is moving slowly or at high speed, while requiring only multiplication operations to compute.
Solution Approach 2:
The patent makes the reference trajectory dynamic by including velocity as a time-varying parameter. This allows the trajectory to adapt to changing driving conditions in real-time, improving accuracy and comfort without requiring complex predictive models or iterative solutions, as velocity is directly measurable from sensors.
4Productivity
If the reference trajectory does not vary over time, then the computational load is lower, but the lateral control performance and collision avoidance capability worsen
Solution Approach 1:
The patent introduces time-varying dynamics into the reference trajectory to improve lateral control performance and collision avoidance. The trajectory evolves over the prediction horizon based on the vehicle's current state, enabling the controller to anticipate future positions and adjust steering commands proactively, thereby enhancing safety without excessive computational burden.
Solution Approach 2:
The patent uses parameter changes over time to improve reliability. The reference trajectory parameters are updated continuously based on measured vehicle state (lateral offset, arc-length, velocity), allowing the control system to adapt to changing conditions and maintain accurate lane centering, which directly improves lateral control performance and collision avoidance capability.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A computer implemented method for trajectory planning of a vehicle comprises the following steps carried out by computer hardware components: determining a reference for a trajectory of the vehicle for a pre-determined prediction time horizon, wherein the reference varies over time during the prediction time horizon; determining a location error based on the reference for the trajectory; and determining a steering command for the vehicle based on the location error.