Machine Learning Component Selection for Circuit Characteristic Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional motor assembly equipment relies on manual component selection, leading to variability in product characteristics and inefficiencies due to random component assembly, requiring significant time and effort.
Innovation Solution
Integration of a machine learning system with a state observer, reward calculator, artificial intelligence, and decision maker to automate the selection of components based on characteristic values, optimizing product assembly and inventory management to minimize variation in circuit characteristics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If manual component selection is used, then flexibility in component choice is maintained, but time consumption and labor effort increase significantly
Solution Approach 1:
The patent replaces the manual mechanical selection process with an automated computer-based system that uses characteristic value data to automatically select components. The system reads characteristic values from storage, calculates compatibility, and automatically selects components without manual intervention, thereby reducing time consumption while maintaining selection flexibility through algorithmic decision-making.
2Ease of manufacture
If random component assembly is used, then assembly simplicity is maintained, but product characteristic variation increases
Solution Approach 1:
The patent performs preliminary sorting and grouping of components based on their characteristic values before assembly. Components are pre-categorized into groups with similar characteristics, and the system selectively chooses components from specific groups to ensure consistent product characteristics. This preliminary classification enables controlled assembly without compromising simplicity.
3Manufacturing precision
If components are selected to optimize product characteristics, then product quality improves, but inventory management complexity increases
Solution Approach 1:
The patent manages inventory complexity by organizing components based on their characteristic value parameters. Components are grouped into categories (first through fifth groups) based on specific parameter ranges, allowing the system to control product characteristics by selecting from predefined groups. This parameter-based organization simplifies inventory management while maintaining precise control over product quality.
Data Source
AI summary
Production equipment according to an embodiment of the present invention includes a machine learning system and an assembly and test unit. The assembly and test unit chooses components from component groups having different characteristic values, assembles the chosen components into a product, and tests the assembled product. The machine learning system includes a state observer for observing a test result of the product and the inventory amounts of the components grouped based on the characteristic values of the components; a reward calculator for calculating a reward based on the test result and the inventory amounts; an artificial intelligence for determining an action value based on an observation result by the state observer and the reward calculated by the reward calculator; and a decision maker for choosing components to be used in the next product assembly from the component groups based on a determination result by the artificial intelligence.