Machine Tool Task Scheduling With Hierarchical Dependency Inference
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional machine tools face challenges in generating optimal multi-task learning models due to equipment heterogeneity, data imbalance, and lack of consideration for task dependencies, leading to reduced reliability, increased maintenance costs, and safety issues.
Innovation Solution
A multi-task real-time inference scheduling system where a central control unit connects to individual control units via a network, collects and processes use contexts through a neural network to generate multi-task learning models, inferring and scheduling tasks based on real-time scenarios, considering dependencies and environments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional multi-tasking by machine learning through general neural networks is used, then machine tools can perform multiple tasks, but data imbalance and lack of task dependency consideration lead to reduced reliability
Solution Approach 1:
The patent segments the multi-task learning system into hierarchical levels (L0-L3) where each level handles specific task types. This segmentation allows balanced data distribution across different task categories and enables the system to handle task dependencies systematically, resolving the reliability issue caused by data imbalance and lack of task dependency consideration.
Solution Approach 2:
The patent introduces task dependency graphs and hierarchical level structures as intermediaries between raw data and the neural network processing. These intermediaries organize tasks based on their dependencies and relationships, allowing the system to process tasks in the correct sequence and consider task relationships, thereby improving reliability.
2Productivity
If machine tools perform machining of diverse and complex shapes with rapid and precise machining, then productivity increases, but equipment heterogeneity and data acquisition uniformity issues arise
Solution Approach 1:
The patent creates a universal hierarchical multi-task learning framework that can handle diverse machining tasks (L0-L3 levels) across different machine tool configurations. This universal framework reduces system complexity by providing a standardized approach to handle various machining operations, workpiece types, and equipment variations through a single integrated system.
Solution Approach 2:
The patent changes the organizational parameters of task processing by introducing hierarchical levels and dependency relationships. Instead of treating all tasks uniformly, the system organizes tasks into levels (L0-L3) with specific parameter ranges and dependencies, allowing efficient handling of diverse machining operations while managing complexity through structured parameter organization.
3Ease of operation
If fixed operations without considering multi-tasking priorities are used, then system simplicity is maintained, but fault diagnosis and monitoring reliability decrease
Solution Approach 1:
The patent performs preliminary organization of tasks into hierarchical levels and establishes task dependency relationships before executing the multi-task learning process. This preliminary structuring of tasks based on their dependencies and priorities enables reliable fault diagnosis and monitoring while maintaining ease of operation, as the system automatically follows the pre-established task hierarchy without requiring complex real-time decision-making.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
The present invention relates to a multi-task real-time inference scheduling system and real-time inference scheduling method of a machine tool, wherein a central control unit is connected to each of one or more individual control units through a network, receives a use context of each machine tool through each individual control unit, generates a multi-task learning model through a neural network, infers multiple tasks required to be performed by the individual control unit of each machine tool through machine learning by using real-time use contexts collected during operation of the machine tool by a use scenario, and schedules the multiple tasks of the machine tool through machine learning.