Continuous Flat Pressing of MDF Quality Control Model Framework and Collaborative Programming Approach Based on Wood Fiber Hot Pressing Mechanism

Yunlei Lv, Yaqiu Liu, Weipeng JingFrontiers in Plant Science 2022 · DOI

TL;DR

This paper introduces a novel quality control framework for Medium-Density Fiberboard (MDF) production that treats the continuous hot-pressing array as a "cellular" system. By combining digital twin modeling with a distributed coordination approach, the researchers achieved a quality superiority rate of over 95%, significantly improving the response time and precision of slab formation.

What Problem Does This Paper Address?

Producing MDF is a high-stakes balancing act. During the continuous hot-pressing stage, wood fibers are subjected to a "multi-field coupling" effect—a complex interplay of heat, moisture, and pressure. Because wood is a viscoelastic material (meaning it behaves somewhere between a liquid and a solid when squeezed), it doesn't always stay the shape you press it into.

Traditional control systems often struggle with the "uncertainty interference" caused by these internal dynamics, leading to common defects like thickness deviations, slopes, depressions, or bulges in the final board. The authors identify a gap in existing research: a lack of a systematic, mechanism-based framework that can coordinate the dozens of hydraulic cylinders in a modern press to react to these deviations in real-time.

Methods

The researchers developed a multi-layered control strategy based on several sophisticated concepts:

  • Cellular Automata (CA) Theory: They treated the array of hydraulic cylinders (the "Continuous Panel System" or CPS) as a grid of individual cells. This allows for localized control that contributes to a global objective.
  • Digital Twin Modeling: A conceptual digital twin was built to map the physical pressing mechanism into a virtual space, allowing for precise sequence parameter control.
  • The Burgers Model: To account for the "squishy" nature of wood fibers, they used the Burgers four-element model to simulate viscoelasticity, ensuring the control system understands how the slab will rebound after pressure is applied.
  • Distributed Coordination Framework: They defined five basic "mode elements" (Modes 0–4) and five "mutual synergy frameworks" (Modes A–E). These range from independent cylinder control to complex "master-slave" relationships where one cylinder leads and others follow to maintain a specific surface slope or thickness.
  • Expert Rule-Based Regulation: The system identifies four specific deviation types (thickness, slope, depression, bulge) and applies a pre-programmed "expert" sequence of pressure and distance adjustments to correct them on the fly.

Datasets

The study did not use a traditional static machine learning dataset. Instead, it utilized real-time production data from an industrial MDF production line with an annual capacity of 300,000 $m^3$. The framework was specifically tested and validated using production specifications for 5.5 mm and 18.5 mm MDF slabs.

Key Insights and Ideas

  • The Press as a "Cellular" Array: Instead of viewing the press as one giant machine, treating it as a coordinated swarm of "cellular" cylinders allows for much finer control over the board's topography.
  • Viscoelasticity is Non-Negotiable: You cannot control MDF quality without math that accounts for the material's internal "memory" and rebound. Integrating the Burgers model directly into the control law was a key differentiator.
  • Master-Slave Synergy: The most effective way to handle complex deviations (like a "bulge" in the middle of the board) is not to adjust every cylinder equally, but to establish a hierarchy where specific cylinders act as anchors (masters) and others adjust relative to them (slaves).
  • Multi-Zone Programming: The press is divided into functional zones (rapid closing, gap keeping, and thickness normalization). The paper shows that control logic must shift fundamentally as the slab moves through these different physical states.

Key Results

The application tests on a live production line yielded impressive metrics:

  • Quality Superiority Rate: Controlled consistently above 95%.
  • Response Time: The distributed coordination approach significantly reduced the time required for the press to adjust to detected deviations compared to traditional methods.
  • Precision: The system effectively counteracted multi-field coupling interference, maintaining tight tolerances even for thin (5.5 mm) boards which are typically more prone to warping and thickness errors.

Limitations and Future Work

While the framework is highly effective at identifying and correcting the type of deviation, the authors note that the intensity and classification levels of these deviations need more granular research. Currently, the "expert rules" handle typical cases well, but a more refined regulation sequence that scales precisely with the magnitude of the error is the next logical step. Additionally, the authors suggest further expanding the control programming to handle even more complex interactions between different levels of deviation types.

Summary updated

Abstract

The increasing demand for forestry resources is driving the need for smarter systems capable of saving and protecting forests that can optimize agile forestry production. This study uses the continuous hot-pressing process of wooden medium-density fiberboard (MDF) to investigate the possibility of automatic quality control of the continuous flat pressing process. For this purpose, conceptual digital twin modeling for mechanism and sequence parameter control was conducted based on the cellular automata (CA) theory. A distributed coordination mode framework was constructed, and a craft control programming method was proposed for the quality control of MDF continuous flat pressing. Based on the MDF continuous flat press craft mechanism and control standards, a framework of five distributed flat press cooperative control mode elements for the cylinder array of the continuous panel system (CPS) was defined. To satisfy the distributed distance servo and pressure servo demands of the multi-stage hot pressing craft design, five kinds of synergy collaborative control modes of multiple rack groups were constructed using mode elements: For the four types of typical deviations in slab production, i.e., thickness, slope, depression, and bulge, a multi-zone mutual cooperative mode craft control sequence was programmed. According to the type and intensity of real-time deviation, the corresponding regulation sequence was applied. This effectively counteracts the deviation caused by the uncertainty interference due to the multi-field coupling effect in actual production. The application tests demonstrate that the adjustment and response time of the continuous flat press were greatly improved, and the quality superiority rate is controlled above 95%, thereby confirming the effectiveness of the control strategy.
BibTeX
@inproceedings{lv2022continuous,
  author = {Yunlei Lv and Yaqiu Liu and Weipeng Jing},
  title = {{Continuous Flat Pressing of MDF Quality Control Model Framework and Collaborative Programming Approach Based on Wood Fiber Hot Pressing Mechanism}},
  booktitle = {Frontiers in Plant Science},
  year = {2022},
  doi = {10.3389/fpls.2022.851219},
}
Figure