Collective Vehicle Maneuver Planning Using Formation Trees
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems struggle to efficiently determine collective maneuvers of autonomously driving vehicles, particularly in complex traffic scenarios involving multiple vehicles and obstacles, due to the uncertainty in predicting the movements of surrounding vehicles and the need for precise, deterministic movement planning.
Innovation Solution
A method that involves determining a current formation of vehicles and obstacles using a local Frenét coordinate system, generating a tree data structure to iteratively expand possible maneuvers, and calculating cost-optimized trajectories to ensure collision-free collective maneuvers, allowing the ego vehicle to adapt flexibly to changing traffic scenarios.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual determination of possible maneuvers by manufacturer is used, then the system is simple to implement, but it cannot efficiently determine collective maneuvers of multiple vehicles in complex traffic scenarios
Solution Approach 1:
The maneuver determination process is segmented into distinct modules: receiving state data from multiple vehicles, determining current formation, generating possible collective maneuvers, calculating trajectories, and selecting optimal maneuvers. This segmentation enables efficient processing of complex multi-vehicle scenarios while maintaining manageable system complexity through modular architecture.
Solution Approach 2:
The system dynamically adapts the set of possible maneuvers based on the current formation and relative positions of vehicles. Rather than using fixed manufacturer-defined maneuvers, the system generates maneuvers in real-time based on dynamic traffic conditions, improving efficiency in complex scenarios while using algorithms of controlled complexity.
2Adaptability or versatility
If the system determines collective maneuvers for all vehicle combinations, then adaptability to traffic scenarios is improved, but computational complexity increases
Solution Approach 1:
The system applies different processing levels to different vehicle combinations based on their relevance to the ego vehicle. Full collective maneuver determination is applied to directly interacting vehicles, while less intensive processing is applied to distant or less relevant vehicles. This local differentiation improves adaptability to relevant traffic scenarios while controlling overall computational complexity.
Solution Approach 2:
The system determines maneuvers for a subset of vehicle combinations that are most relevant to the ego vehicle's path and safety, rather than exhaustively analyzing all possible combinations. This partial action approach provides sufficient adaptability for safe operation while significantly reducing computational complexity compared to complete enumeration of all vehicle interactions.
3Measurement precision
If trajectories are calculated for all possible collective maneuvers, then maneuver accuracy is improved, but calculation time increases
Solution Approach 1:
The system pre-calculates and stores the set of possible collective maneuvers and their associated trajectories before actual maneuver execution is needed. This preliminary action allows the system to have accurate trajectory data ready when needed, improving precision while reducing real-time calculation time by performing computations in advance when computational resources are more abundant.
Data Source
AI summary
A method for determining a collective maneuver of at least two vehicles includes receiving a state of a first vehicle and a state of at least one further vehicle in an environment of an ego vehicle, and determining a current formation for the first vehicle and the at least one further vehicle on the basis of the received state of the first vehicle and the received state of the at least one further vehicle in the environment of the ego vehicle, the current formation comprising a relative arrangement of the first vehicle and of the at least one further vehicle with respect to a roadway. The method also includes determining a set of collective maneuvers on the basis of the current formation, where each collective maneuver of the set of collective maneuvers comprising a sequence of formations from the current formation to an end formation. A trajectory for the first vehicle and a trajectory for the at least one further vehicle is calculated for a collective maneuver from the set of collective maneuvers, and the collective maneuver from the set of collective maneuvers is determined based on the calculated trajectories of the collective maneuver from the set of collective maneuvers and a motion of the first vehicle and of the at least one further vehicle sensed by means of a sensor system of the ego vehicle.