Fleet Allocation via Branch and Bound MCTS
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Selecting the optimal fleet composition of autonomous mobile robots for material delivery in automated facilities is a computationally intensive problem that requires significant resources, making it challenging to minimize costs effectively.
Innovation Solution
A method integrating the Branch and Bound algorithm with Monte Carlo Tree Search to determine the sequence of waypoints, path planning using dual graphs, and fleet composition, which partitions the search space and updates upper bounds to find the globally optimal fleet composition efficiently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional optimization methods are used to find optimal fleet composition, then global optimal solution is achieved, but computational resources and time required become excessively large
Solution Approach 1:
The patent segments the complex fleet composition optimization problem into multiple sub-problems by dividing the search space into different levels (e.g., number of robots, robot types, task assignments). This allows the Branch and Bound algorithm to systematically explore and prune the search space, finding optimal solutions without exhaustively checking all possibilities, thus reducing computation time while maintaining optimality.
Solution Approach 2:
The patent employs preliminary actions by using the Branch and Bound algorithm to pre-determine upper bounds and prune the search space before the main optimization process. This preliminary structuring of the search space allows subsequent algorithms to work more efficiently, reducing the overall computational burden while ensuring optimal fleet composition is found.
2Loss of energy
If comprehensive fleet composition optimization is performed, then optimal cost minimization is achieved, but computational complexity increases significantly
Solution Approach 1:
The patent segments the optimization problem into hierarchical levels: first determining the optimal number of robots, then the mix of robot types, and finally the task assignments. Each level has its own search space that can be independently optimized using Branch and Bound, reducing the computational complexity of the overall problem while minimizing operational costs through systematic exploration of all feasible combinations.
Solution Approach 2:
The patent introduces intermediary algorithms and data structures (such as the Branch and Bound framework with upper bound tracking) that mediate between the raw problem data and the final optimization. These intermediaries organize the computational tasks, manage the search space, and provide pruning criteria, thereby reducing the effective computational complexity while maintaining comprehensive cost optimization.
3Reliability
If all constraints of factory operation are considered in fleet selection, then realistic optimal solution is achieved, but the problem size becomes exceedingly large
Solution Approach 1:
The patent segments the search space by organizing constraints into different categories (robot capabilities, task requirements, operational constraints) and addressing them in hierarchical levels. This segmentation allows the Branch and Bound algorithm to systematically incorporate all constraints without exploding the search space, as each constraint level can be processed independently and pruned appropriately.
Solution Approach 2:
The patent performs preliminary filtering and classification of constraints before the main optimization process. By pre-organizing and pre-processing the constraint information, the algorithm can efficiently navigate the search space while respecting all factory operational constraints, reducing the effective search space size without sacrificing the realism of the solution.
Data Source
AI summary
A method of managing a fleet of robots for delivery of materials in a facility is provided. The method includes: determining a sequence of waypoints by a Branch and Bound (B&B) method; determining a path through the sequence of waypoints by a dual graph method; and determining a fleet composition and distribution of tasks among the robots by the B&B method and a Monte Carlo Tree Search (MCTS) method.