CNC Abnormality Detection Using Condition-Specific Learning Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing numerical control systems face challenges in detecting abnormalities in machine tool operation states due to varying machining conditions, such as motor operation patterns, tool types, and workpiece materials, making it difficult to create a general-purpose machine learning device that can effectively detect abnormalities across different situations.
Innovation Solution
A numerical control system that switches learning models based on specific operation conditions, including motor operation patterns, tool types, and workpiece materials, allowing for machine learning and abnormality detection tailored to these conditions, using a condition designating unit, state amount detection unit, inference computing unit, abnormality detection unit, learning model generation unit, and learning model storage unit.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a general-purpose machine learning device is created to detect abnormalities across various machining situations, then the detection coverage is improved, but the device complexity and difficulty of creation increase significantly
Solution Approach 1:
The patent divides the general-purpose abnormality detection system into multiple specialized learning models, each trained for specific machining conditions (e.g., rough machining, finish machining, different workpiece materials). This segmentation allows the system to maintain high detection accuracy for each condition while avoiding the complexity of a single universal model that must handle all scenarios.
Solution Approach 2:
The system dynamically selects and switches between different learning models based on the current machining conditions. The learning model selection unit determines which model to use based on parameters such as machining type, workpiece material, and tool type, allowing the system to adapt to varying conditions without requiring a single complex model to handle all scenarios.
2Device complexity
If machine learning is performed with a single general-purpose model, then the system structure is simplified, but the detection accuracy decreases due to over-learning and inability to specialize in specific conditions
Solution Approach 1:
Each learning model is specialized for specific local conditions (particular machining operations, materials, or tools), allowing it to achieve high detection accuracy for those specific scenarios. The system applies the appropriate local model based on current conditions, ensuring optimal detection performance without requiring a single complex general-purpose model.
Solution Approach 2:
The system changes the parameters of the learning model being used based on machining conditions. By selecting different models trained on different parameter sets (e.g., different workpiece materials, tool types, machining operations), the system maintains high detection accuracy across varying conditions while keeping each individual model relatively simple.
3Measurement precision
If multiple learning models are maintained for different machining conditions, then the detection accuracy for specific conditions is improved, but the device complexity and model management burden increase
Solution Approach 1:
The learning model selection unit serves as a universal interface that manages multiple specialized models. It automatically determines which model to apply based on current machining conditions, providing a unified control mechanism that simplifies the management of multiple models while maintaining high detection accuracy for each specific condition.
Solution Approach 2:
The system uses feedback from the machining condition parameters (workpiece material, tool type, machining operation) to select the appropriate learning model. This feedback mechanism automates the model selection process, reducing the management burden while ensuring the most accurate model is used for each specific machining scenario.
4Ease of manufacture
If a single learning model is used for all machining operations, then the system is easier to implement, but it fails to detect abnormalities accurately when machining conditions vary
Solution Approach 1:
The system dynamically adapts to varying machining conditions by selecting the appropriate pre-trained learning model for each scenario. This dynamic model selection maintains implementation simplicity (models are pre-trained and stored) while significantly improving detection reliability by using condition-specific models rather than a single general-purpose model.
Data Source
AI summary
A numerical control system detects a state amount indicating an operation state of a machine tool, creates a characteristic amount that characterizes the state of a machining operation from the detected state amount, infers an evaluation value of the operation state of the machine tool from the characteristic amount, and detects an abnormality in the operation state of the machine tool on the basis of the inferred evaluation value. The numerical control system generates and updates a learning model by machine learning that uses the characteristic amount, and stores the learning model in correlation with a combination of conditions of the machining operation of the machine tool.