AGV Learning Device Optimizing Production Quantity and Stock
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for optimizing AGV route planning in production sites primarily focus on physical indicators, such as minimizing transportation time, but fail to consider high-level logical indicators like production quantity and stock management, which are crucial for overall profitability.
Innovation Solution
A learning device and system that uses a simulator to generate training data incorporating high-level indicators like production quantity and stock management, along with a reward function to learn a value function for deriving an optimal policy for AGVs, optimizing both physical and logical indicators.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If shortest route planning is used to minimize transportation time, then physical indicators are optimized, but high-level production indicators are not necessarily satisfied
Solution Approach 1:
The patent combines multiple previously separate planning processes (shortest route planning, stock control planning, production quantity planning) into a single integrated planning process. The planning unit simultaneously considers physical indicators (transportation time) and high-level production indicators (production quantity, stock quantity) to generate unified plans that optimize overall factory performance rather than individual metrics in isolation.
Solution Approach 2:
The planning unit is designed to perform multiple functions simultaneously: it conducts shortest route planning for AGVs, stock control planning for inventory management, and production quantity planning for production optimization. This multi-functional planning system can adapt to different planning scenarios and optimize various indicators based on the specific situation, making the system universally applicable to different production needs.
2Speed
If focus is placed on physical indicators like transportation time, then route efficiency is improved, but logical indicators like stock management are neglected
Solution Approach 1:
The patent merges the previously separate planning processes (shortest route planning, stock control planning, production quantity planning) into a unified planning system. The planning unit simultaneously processes physical indicators (transportation time, routes) and logical indicators (stock quantity, production quantity) to generate integrated plans that optimize both aspects together, preventing the loss or neglect of production indicator information.
3Adaptability or versatility
If multiple separate planning processes are used for different indicators, then each indicator can be optimized independently, but overall coordination and profitability are reduced
Solution Approach 1:
The patent merges multiple separate planning processes into a single integrated planning unit that simultaneously handles shortest route planning, stock control planning, and production quantity planning. This unified approach maintains the adaptability to handle different planning scenarios while improving overall factory performance through coordinated optimization of all indicators together rather than independent optimization that lacks coordination.
Data Source
AI summary
The input means 81 accepts input of a reward function that defines cumulative reward by a reward term based on a high-level indicator representing a production indicator. The learning means 82 learns a value function for deriving optimal policy for an agent using training data and the reward function. The output means 83 outputs the learned value function.