Fibre Moisture Prediction Control for Stable Fibreboard Drying
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current control systems for fibreboard manufacturing processes fail to accurately predict and adjust moisture content at the end of the drying stage, leading to potential low-quality products due to high or low moisture levels, as they do not consider all relevant manufacturing parameters from debarking to the final cyclone, and are reactive rather than proactive.
Innovation Solution
A neural network-based moisture control system that uses real-time data from sensors across the production line to predict fibre moisture content at the drying stage and adjust input drying temperatures to meet setpoints, incorporating parameters from previous stages like defibrating and digestion, to optimize moisture levels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If current control systems are used to monitor moisture content at the drying stage output, then moisture content can be measured, but the system cannot accurately predict or adjust moisture content because it does not consider all relevant manufacturing parameters from previous stages
Solution Approach 1:
The neural network predicts the moisture content at the drying stage output in advance, before the actual drying process completes. This allows the system to anticipate the final moisture content and adjust drying parameters proactively, rather than merely measuring after the fact. The prediction is based on parameters from previous stages (defibrating, digestion, mixing) to enable early intervention.
Solution Approach 2:
The system implements a closed-loop control mechanism where the neural network's prediction of moisture content feeds back to adjust the drying parameters. The control system continuously monitors parameters from previous stages, updates the prediction, and modifies drying conditions accordingly, creating a dynamic feedback loop that improves both measurement precision and control reliability.
2Ease of manufacture
If reactive control systems are used to adjust drying parameters after moisture deviation occurs, then some correction can be made, but the system cannot prevent quality issues because it lacks predictive capability
Solution Approach 1:
The neural network performs preliminary prediction of moisture content based on parameters from previous manufacturing stages. This allows the system to identify potential moisture deviations before they manifest in the final product, enabling preventive adjustment of drying parameters rather than reactive correction after quality issues arise.
Solution Approach 2:
The system applies preliminary anti-action by predicting moisture content deviations and counteracting them in advance through parameter adjustment. Instead of allowing moisture content to deviate and then correcting it, the neural network anticipates the deviation and pre-adjusts drying parameters to prevent the quality issue from occurring in the first place.
3Device complexity
If drying parameters are adjusted without considering parameters from previous stages like defibrating and digestion, then the control system is simpler, but the prediction accuracy of moisture content deteriorates
Solution Approach 1:
The neural network serves multiple functions: it monitors parameters from previous stages (defibrating, digestion, mixing), predicts moisture content at the drying stage output, and triggers parameter adjustments. This multi-functional approach consolidates what could be multiple separate systems into a single intelligent controller, managing complexity while improving prediction accuracy through comprehensive parameter consideration.
Solution Approach 2:
The system dynamically changes parameters based on their relevance to moisture content prediction. The neural network identifies which parameters from previous stages have the greatest impact on final moisture content and focuses computational resources on those critical parameters, rather than uniformly processing all possible parameters, thus balancing complexity and accuracy.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach allows for proactive adjustment of drying process parameters, ensuring consistent fibre moisture content and improving the quality of fibreboards by anticipating and minimizing deviations from set moisture levels, thus enhancing the efficiency and homogeneity of the fibreboard production process.
Implementation Method 1
A moisture control unit is provided which is configured to receive a plurality of inputs corresponding to parameters related to the fibreboard manufacturing process, to generate, using a neural network that comprises at least one hidden neural network layer, an estimate or prediction of the fibre moisture content at the output of the drying stage based on the received inputs
Implementation Method 2
the first drying step may comprise pushing the fibre up to the top of a cyclone with hot air generated by different boilers, generators, or similar. This first drying step removes most of the moisture from the fibres. Upon entering the cyclone, the fibre precipitates by gravity while the moisture removed from the fibres is expelled outside by natural convection
Implementation Method 3
Once the fibre has been decanted in the first cyclone through a cyclonic centrifugation process
Implementation Method 4
the fibre precipitates by gravity while the moisture removed from the fibres is expelled outside
Implementation Method 5
The moisture sensors generally determine the volumetric water content in the chips by measuring some property of the wood, such as electrical resistance, dielectric constant or by measuring reflection of emitted electromagnetic signals
Implementation Method 6
The moisture sensors generally determine the volumetric water content in the chips by measuring some property of the wood, such as electrical resistance, dielectric constant or by measuring reflection of emitted electromagnetic signals
Implementation Method 7
Prior to the introduction of chips into the digester, the wood chips may be transported to a hopper where they are preheated with, for example, steam (e.g., at 80-90°C). This pre-heating stage reduces the hardness of the chips such they are easy to squeeze by the squeezing screw
Implementation Method 8
Then, the chips can be cooked in the digester by adding steam
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Examples refer to a moisture content control system and method for controlling a fibre moisture content in a fibreboard manufacturing process. The system comprises a plurality of sensors to monitor parameters of the fibreboard manufacturing process, a plurality of feedback loop controllers to adjust parameters of the fibreboard manufacturing process based on differences between pre-defined setpoints associated to the parameters and current value of said parameters and a moisture control unit. The moisture control units receives a plurality of inputs, generates, using a neural network, a prediction of the fibre moisture content at the output of the drying stage, compares the prediction with a pre-defined setpoint of the fibre moisture content at the output of the drying stage and modifies at least one setpoint associated to a corresponding input drying temperature of a respective drying unit of the fibreboard manufacturing process based on the result of the comparison.