Paper Machine Web-Break Prediction With Adaptive Auto-Labeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for predicting web breaks in paper machines are inefficient due to cumbersome manual data labeling and the use of unsupervised models with unlabeled data, leading to unreliable predictions and increased downtime.
Innovation Solution
A method using machine learning models to automatically label simulated and historical parameters as normal or abnormal patterns, selecting the best model based on performance metrics, and using it to predict web breaks, determine root causes, and estimate time to failure, with adaptive learning to improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual data labeling is used, then labeling accuracy can be maintained, but the process becomes cumbersome and infeasible as the number of parameters increases
Solution Approach 1:
The system uses unsupervised learning models to automatically label data without human intervention. The model learns patterns from unlabeled data and autonomously assigns labels to normal and abnormal patterns, eliminating the need for manual labeling while maintaining accuracy through automated pattern recognition and classification
Solution Approach 2:
The patent replaces the mechanical manual labeling process with an automated machine learning-based labeling system. The unsupervised model uses computational algorithms to analyze patterns in process parameters and automatically generate labels, substituting human expertise with an automated intelligent system
2Extent of automation
If unsupervised models are used with unlabeled data, then manual labeling effort is reduced, but prediction robustness deteriorates
Solution Approach 1:
The system incorporates feedback mechanisms where the unsupervised model continuously learns from labeled data and refines its predictions. The model uses feedback from pattern recognition to improve its labeling accuracy over time, ensuring robust predictions while maintaining automated operation
Solution Approach 2:
The patent changes the approach by using multiple process parameters and operational parameters together to improve prediction robustness. The system analyzes changes in parameter patterns over time and uses these parameter transformations to enhance the reliability of break predictions while maintaining automated labeling
3Measurement precision
If more parameters are monitored to improve prediction accuracy, then detection precision improves, but system complexity increases
Solution Approach 1:
The system segments the monitoring process by categorizing parameters into different types (process parameters and operational parameters) and analyzing them separately through the unsupervised model. This segmentation allows the system to handle multiple parameters systematically without overwhelming complexity, improving detection precision through structured analysis
Data Source
AI summary
A system and method for labelling normal and abnormal regions in data related to a paper machine for web break prediction and labelling individual parameters for root cause analysis, using machine learning models, includes using the machine learning models in real-time to predict breaks in the paper web, analyzing root cause for the breaks in the paper web, and estimating a time to break. An auto-data-labeling framework helps in adaptive learning for autonomous model improvement of the deployed model, transfer learning, shortlisting parameters and automating feasibility study.