Injection Molding Check Valve Wear Detection With Switched AI Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting wear in check valves of injection molding machines are inefficient, particularly when operation conditions and environmental conditions vary, leading to difficulties in accurately determining wear states and potentially causing resin backflow and inconsistent product weights.
Innovation Solution
A numerical control system that switches learning models based on operation and environmental conditions to detect wear in check valves, using a condition designating unit, state amount detection, inference computing, abnormality detection, and learning model generation to select and generate learning models for accurate wear detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a general-purpose machine learning device is introduced to estimate wear state, then wear detection capability is improved, but over-learning occurs and accuracy decreases due to excessive data requirements
Solution Approach 1:
The patent segments the general wear detection problem into multiple condition-specific sub-problems by dividing learning models into condition-specific models. Each model is trained only on data relevant to its specific condition (e.g., specific resin type, screw diameter, injection molding machine size), preventing over-learning while maintaining comprehensive coverage across all operating scenarios.
Solution Approach 2:
The system dynamically selects the appropriate learning model based on current operating conditions. The control device determines which condition-specific model to use by matching current parameters (resin viscosity, screw diameter, machine size) with the conditions for which each model was trained, ensuring optimal accuracy for the current situation.
2Device complexity
If conventional wear detection methods are used, then simplicity is maintained, but accuracy decreases when operation conditions vary
Solution Approach 1:
The patent creates a universal wear detection system that handles multiple operating conditions through condition-specific learning models. Each model is specialized for its condition, yet the overall system is universal in covering all possible operating scenarios. The control device automatically selects the appropriate model based on current conditions, maintaining simplicity while achieving high accuracy across variations.
3Measurement precision
If direct measurement of check valve dimension is performed, then measurement accuracy is improved, but productivity decreases due to production stoppage
Solution Approach 1:
The patent replaces direct mechanical measurement of the check valve with an indirect detection system using machine learning. Instead of physically removing and measuring the valve, the system uses sensors to detect operating parameters (torque, pressure, temperature) and uses condition-specific learning models to infer wear state, eliminating production stoppage while maintaining detection capability.
Solution Approach 2:
The learning model acts as an intermediary between raw sensor data and wear state determination. Rather than directly measuring the check valve, the system uses multiple sensor readings as intermediaries to indirectly assess valve condition, enabling continuous monitoring without interrupting the molding process.
Data Source
AI summary
A numerical control system detects a state amount indicating a state of an injection operation of an injection molding machine, generates a characteristic amount that characterizes the state of the injection operation from the state amount, and infers an evaluation value of the state of the injection operation from the characteristic amount. The numerical control system detects an abnormal state on the basis of the evaluation value, generates or 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 injection operation.