Autonomous Vehicle Trajectory Optimization via Bending Band Model
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current autonomous driving methods do not effectively optimize vehicle trajectories for desired driving properties, such as comfort and energy efficiency, especially in situations that require adaptation to changing road conditions and vehicle characteristics.
Innovation Solution
A method using a bending band defined by discrete elements, where parameters like flexural and torsional rigidity are optimized using the finite element method and principle of virtual displacements to adapt trajectory properties situationally, allowing for local optimization of vehicle motion and reduced computational complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If full-distance trajectory planning is performed with high computational precision, then trajectory optimization quality is improved, but computational expenditure increases significantly
Solution Approach 1:
The trajectory planning is divided into multiple discrete elements along the bending band. Each element can be independently optimized with local parameters, allowing the system to achieve high overall optimization quality without computing every detail at full resolution simultaneously. This segmentation enables efficient computational resource allocation.
Solution Approach 2:
Different sections of the trajectory are assigned different optimization priorities and parameter resolutions based on their importance and complexity. Critical sections receive higher computational precision while less critical sections use coarser optimization, thereby maintaining overall trajectory quality while reducing total computational expenditure.
2Measurement precision
If the number of discrete elements in the bending band is doubled, then trajectory precision is improved, but computational expenditure increases eight fold
Solution Approach 1:
The system dynamically adjusts the number and density of discrete elements based on the specific driving situation, road geometry, and optimization requirements. Rather than using a fixed high number of elements always, the system adapts the discretization level to match the actual complexity needed, preventing unnecessary computational overhead while maintaining required precision.
Solution Approach 2:
The optimization algorithm changes parameters such as element density, refinement level, and calculation precision adaptively during the planning process. This allows the system to achieve high trajectory precision when needed while using coarser approximations when acceptable, thereby avoiding the eight-fold computational increase that would result from consistently using maximum precision.
3Ease of operation
If trajectory optimization considers multiple optimization goals and parameters, then driving comfort and energy efficiency are improved, but device complexity increases
Solution Approach 1:
Multiple optimization goals such as driving comfort, energy efficiency, and trajectory accuracy are merged into a unified optimization framework using the bending band model. This integrated approach allows the system to simultaneously consider multiple parameters without requiring separate complex control systems for each goal, thereby improving overall performance while managing system complexity.
Solution Approach 2:
The bending band optimization system serves multiple functions simultaneously: it optimizes trajectory geometry, adjusts for driving comfort, reduces energy consumption, and adapts to different road conditions. This multi-functional approach eliminates the need for separate specialized systems for each optimization goal, reducing overall device complexity while achieving comprehensive optimization.
Data Source
AI summary
Method and device for optimized autonomous driving, wherein a trajectory for the actuation of the vehicle is determined. A profile of the trajectory is defined by a bending line of a bending band. The bending line is preferably determined by the finite element method, in particular according to the principle of virtual shifting, and achieves an optimization goal and satisfies a boundary condition. The boundary condition is defined in accordance with a profile of a roadway for the vehicle. The optimization goal is defined by a desired driving property of the vehicle.