Autonomous Vehicle Cluster Planning for Shared Resource Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing motion planning algorithms for multiple vehicles face computational inefficiencies due to exponential growth in search space with the number of vehicles, lack of load balancing, and resource utilization conflicts, making them unsuitable for scalable and efficient control.
Innovation Solution
Partition vehicles into clusters based on predefined motion constraints, forming resource subsets to reduce computational effort, and execute tree-based planning algorithms in a cluster-wise manner with optional load balancing and parallelization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If motion planning is performed for all vehicles together, then coordination between vehicles is improved, but computational complexity increases exponentially
Solution Approach 1:
The patent divides the fleet of vehicles into multiple clusters based on spatial regions and motion constraints. Each cluster is planned independently using tree-based algorithms, reducing the computational complexity from exponential growth with total fleet size to manageable levels per cluster. This segmentation maintains coordination within clusters while enabling parallel processing across clusters.
2Productivity
If vehicles are partitioned into spatial regions, then computational tractability is improved, but load balancing deteriorates
Solution Approach 1:
The patent implements dynamic load balancing where the central controller monitors computational workload across different spatial regions and dynamically adjusts cluster assignments. When one region accumulates too many vehicles, the controller can reassign vehicles to other regions or adjust planning parameters to balance the computational load, maintaining system efficiency while preserving spatial partitioning benefits.
3Reliability
If vehicles are reassigned to clusters based on current position, then coordination is improved, but computational overhead increases
Solution Approach 1:
The patent pre-computes and stores motion constraints for each vehicle based on their operational characteristics and environmental factors. These pre-computed constraints are used to determine cluster assignments, eliminating the need for real-time reassignment calculations. This preliminary action reduces computational overhead during execution while maintaining accurate coordination based on vehicle positions and constraints.
Data Source
AI summary
There is provided a method for controlling a plurality of vehicles sharing a set of common resources Each vehicle's motion is controllable by a finite set of predefined commands. The method includes forming subsets of the common resources on the basis of predefined motion constraints of individual ones of the vehicles; partitioning the vehicles into disjoint clusters on the basis of the subsets; executing a tree-based planning algorithm. Each execution including an evaluation of sequences of the predefined commands to be fed to the vehicles in a single cluster and a selection of a preferred sequence of the commands; and feeding the preferred sequence of commands to the vehicles in each cluster.