Multi-Agent Path Adjustment via Utility Function Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing path adjustment systems for multiple moving bodies require administrators to declare large amounts of data and transmit it frequently, leading to inefficiencies due to exponential increases in required combinations and the need to consider space-time prices, making them inconvenient and data-intensive.
Innovation Solution
A path adjustment system that includes a path reception unit for receiving evaluation data, a multi-path optimization unit that uses a learned utility function to optimize paths to prevent collisions, and a transmission unit to send optimized paths, reducing the data administrators need to declare and simplifying path determination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If path adjustment systems use traditional multi-agent pathfinding algorithms with centralized determination, then collision-free paths can be determined, but the system complexity and data transmission requirements increase exponentially
Solution Approach 1:
The patent segments the centralized path determination problem into individual agent-level decisions. Each agent independently determines its path by evaluating utility functions for different paths, rather than a centralized system computing all paths simultaneously. This segmentation reduces system complexity from exponential to polynomial scale while maintaining collision-free determination through utility function comparisons that inherently account for other agents' paths.
Solution Approach 2:
Each agent autonomously determines its own path by evaluating utility functions and selecting the optimal path without requiring centralized coordination. The agent independently assesses path options considering collision avoidance, eliminating the need for complex centralized computation and data transmission while ensuring reliable collision-free path determination through self-service path planning.
2Reliability
If path adjustment systems evaluate all possible path combinations for multiple agents, then optimal collision-free paths are found, but the data transmission requirements increase exponentially
Solution Approach 1:
The patent extracts the essential evaluation criteria from exhaustive path combination analysis and encapsulates them in utility functions. Instead of transmitting and evaluating all possible path combinations, the system extracts only the necessary utility evaluations for individual agent paths, dramatically reducing data transmission volume while maintaining optimal path determination through comparative utility assessment.
Solution Approach 2:
The patent changes the evaluation parameter from exhaustive path combination analysis to utility function-based individual path evaluation. By transforming the problem from evaluating all combinations to evaluating utility functions for each agent's candidate paths, the system reduces data transmission requirements from exponential to manageable levels while preserving optimal path determination through utility comparison.
3Adaptability or versatility
If administrators manually determine paths for multiple moving bodies, then arbitrary utility functions can be applied, but the operation becomes inefficient and data-intensive
Solution Approach 1:
The system enables agents to automatically determine their own paths by evaluating utility functions, eliminating the need for manual administrator intervention. Each agent self-services by autonomously assessing path options based on its utility function and selecting the optimal path, dramatically improving productivity while maintaining the adaptability to use arbitrary utility functions tailored to specific operational requirements.
Solution Approach 2:
The patent transforms path determination from a manual, administrator-driven process to an automated agent-level decision process. By changing the operational parameter from human manual determination to automated utility function evaluation, the system achieves both high productivity through automation and versatility through customizable utility functions that can be configured for different operational scenarios.
Data Source
AI summary
A path reception means 81 receives path evaluation data including each of paths of a plurality of moving bodies and an evaluation value of a business operator for each of the paths. A multi-path optimization means 82 performs optimization based on a utility function that represents adequacy of a business operator for a path learned based on the path evaluation data such that a number of the moving bodies existing in an identical area is a predetermined number or less and determines each of the paths of the plurality of moving bodies to adjust the paths of the plurality of moving bodies. A path transmission means 83 transmits each optimized path.