Cardboard Sheet Warping Control Using Transient-State Prediction Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In cardboard sheet manufacturing, existing systems struggle to promptly suppress warping of cardboard sheets due to variations in control values with production states, leading to delayed feedback control and potential sheet damage.
Innovation Solution
A system that calculates a prediction model excluding transient operation state information, allowing for precise updates to target control values and reducing warping by determining transient states based on temperature, warping, production, and speed changes, and prioritizing worker input for accurate model generation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If feedback control is used to adjust control values after production state changes, then warping suppression accuracy is improved, but response time is delayed
Solution Approach 1:
The system performs preliminary action by calculating prediction models in advance that predict warping amounts for different production states. When a production state change occurs, the system can immediately retrieve and apply the pre-calculated prediction model corresponding to the new state, eliminating the delay inherent in real-time feedback control while maintaining high warping suppression accuracy.
Solution Approach 2:
The system dynamically adapts by maintaining multiple prediction models corresponding to different production states (different paper widths, basis weights, production speeds). The appropriate model is selected and applied based on the current production state, enabling the system to respond dynamically to state changes without the lag of feedback control calculation.
2Manufacturing precision
If control values are adjusted frequently to adapt to production state changes, then warping suppression is improved, but system stability deteriorates
Solution Approach 1:
By pre-calculating prediction models for various production states before actual production begins, the system avoids frequent real-time adjustments. The models are computed in advance when the system is stable, and then applied during production based on the current state, reducing the frequency of control value changes while maintaining effective warping suppression.
Solution Approach 2:
The system changes parameters (control values) based on pre-determined prediction models rather than frequent real-time adjustments. This approach allows the system to adapt to different production states by selecting from pre-optimized parameter sets, maintaining stability while achieving effective warping control through parameter selection rather than continuous modification.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
In a cardboard sheet manufacturing system 100, it is possible to calculate a prediction model with high precision because information used in calculating a prediction model M1 does not include prescribed information to be deleted. As a result, the target control value of the control table can be updated to a precise value, and thus warping of the cardboard sheet can be promptly suppressed.