Method for outputting prediction data for prediction of at least one quality parameter for at least one building material panel

Abstract

There is disclosed, among other things, a method comprising preparing a mathematical model for a production process of at least one building panel, where the mathematical model is based on a number of training data sets corresponding to or greater than a minimum number; wherein for a building panel type, the training data set comprises quality parameter data representing at least one value of at least one corresponding quality characteristic of a building panel sample of the building panel type, and process parameter data representing values of a plurality of process parameters of a production method for producing the building panel sample; wherein the number of training datasets comprises at least one training dataset of a building panel type for the at least one building panel, and wherein the method further comprises obtaining prediction data based on a mathematical model wherein the prediction data represents a prediction of at least one quality parameter for the at least one building panel; and outputting the prediction data.

Description

Technical Field

[0001] Exemplary embodiments of the present invention are directed to methods, apparatus, systems, and computer programs, particularly for outputting prediction data for a prediction of at least one quality parameter for at least one building material panel. Background Art

[0002] The continuous production of material and / or building material panels, such as wood-based panels, usually requires the use of complex production plants, which include a plurality of units or sections provided for the respective production steps. For each of these units, there are usually a large number of process parameters that characterize the respective process conditions. For example, in the production of fiberboards, process parameters such as the amount of crushing water, the amount of sawdust, the cooking temperature, etc. are characteristic for the process conditions in the defibrillation section of a wood-based panel production plant.

[0003] These process parameters have an influence on the quality properties of the building material panels produced under the corresponding production conditions, for example on the transverse tensile strength, bulk density, flexural strength etc.

[0004] In view of the fact that there are usually a large number of process parameters that can characterize the process conditions of a building material panel production plant, it has proven to be advantageous to use mathematical models to describe the relationship between the selected process parameters and the resulting quality characteristics. These models can help calculate the influence of the set and / or measured process parameter target values ​​or actual values ​​on the expected quality characteristics and, in the case of a manufacturing production plant, can also predict the corresponding quality characteristics based on the existing process parameter target values ​​or actual values. Based on the existing process parameter target values ​​or actual values, the corresponding predicted values ​​can be displayed to the operator so that the operator of the production plant can react to the corresponding settings of the production plant and thus appropriately adjust the expected quality characteristics.

[0005] However, it has been found that it is more advantageous if the mathematical model for predicting the quality characteristics is generated specifically for the type of panel to be produced. Since the number of building material panel samples (also called laboratory samples) required to generate the mathematical model for one building material panel type is relatively large, this may result in a relatively long time before a sufficient sample data set is obtained in the case of a building material panel type that accounts for a small proportion of the total production volume before a corresponding mathematical model can be generated for this building material panel type.

[0006] It was also found that once a mathematical model has been developed for a certain type of building material panel, the prediction results obtained by this mathematical model can deteriorate over time, for example depending on seasonal changes. Summary of the invention

[0007] In this context, the present invention is based on the object of providing, in particular, methods, devices, systems and computer programs, in particular for outputting prediction data representing a prediction of at least one quality parameter of at least one building material panel, even if the production of this type of building material panel accounts for a small proportion of the total production. In addition, it is an object of the present invention to provide methods, devices, systems and computer programs that are capable of outputting reliable prediction data, in particular for such building material panel types. Another object of the present invention is to provide methods, devices, systems and computer programs that can support the control of a production plant for manufacturing building material panels.

[0008] According to an exemplary aspect of the present invention, a method is disclosed, for example, performed by at least one device or a system consisting of at least two devices, the method comprising:

[0009] - preparing a mathematical model for the production of at least one building panel, wherein the mathematical model is based on a number of training data sets, the number of training data sets corresponding to the minimum number or being greater than the minimum number;

[0010] - wherein for one building panel type, the training data set comprises quality parameter data representing at least one value of at least one respective quality feature of a building panel sample of the building panel type, and process parameter data representing values ​​of a plurality of process parameters of a production method for producing the building panel sample;

[0011] - wherein the number of training data sets comprises at least one training data set of a building panel type for the at least one building panel, wherein the method further comprises:

[0012] - obtaining prediction data based on the mathematical model, wherein the prediction data represents a prediction of at least one quality parameter for the at least one building panel;

[0013] - Output prediction data.

[0014] According to this aspect of the invention, it is further disclosed that:

[0015] -A computer program, including program instructions, for when the computer program is executed on a processor, it causes the processor to execute and / or control the method according to the aspects of the present invention. In this specification, the processor mainly refers to a control unit, a microprocessor, a microcontroller, such as a microcontroller, a digital signal processor (DSP), an application-specific integrated circuit (ASICs) or a field programmable gate array (FPGAs). Here, all steps of the method can be controlled, all steps of the method can be executed, and one or more steps can be controlled and one or more steps can be executed. The computer program can be at least partially software and / or firmware of the processor. The computer program can also be implemented at least partially in hardware. For example, the computer program can be stored on a computer-readable storage medium, such as a magnetic, electrical, optical and / or other type of storage medium. For example, the storage medium can be a part of the processor, such as a (non-volatile or volatile) program memory of the processor or a part thereof. For example, the storage medium can be a tangible or physical storage medium.

[0016] -A device or a system consisting of at least two devices for executing and / or controlling the method according to the aspect of the present invention, or including respective corresponding configurations for executing and / or controlling the steps of the method according to the aspect of the present invention. Here, all steps of the method can be controlled, all steps of the method can be executed, one or more steps can be controlled and one or more steps can be executed. One or more of the configurations can also be executed and / or controlled by the same unit. For example, one or more of the configurations can be composed of one or more processors. For example, the device according to the aspect of the present invention can be a control device connected to a production device (at least part of it) for manufacturing building material panels, especially for controlling at least part of the method for manufacturing building material panels.

[0017] - A device comprising at least one processor and at least one memory containing program code, wherein the memory and the program code are configured to cause the device with at least one processor to at least execute and / or control the method according to the aspect of the present invention. It can control all steps of the method, execute all steps of the method, or control one or more steps and execute one or more steps.

[0018] Features of the described aspects will be described in part by way of examples below.

[0019] In an exemplary embodiment, the method according to the aspect of the invention may be performed by at least one control device, which is (directly and / or indirectly) connected to a production plant for manufacturing a building material panel, in particular for controlling at least some parts of the method for manufacturing a building material panel. In an exemplary embodiment, the method may be performed by a system, for example, comprising a plurality of control devices, wherein each control device performs (partial or all) steps of the method. For example, the control device may include a processing device, such as one (or more) computers arranged on the production plant (and capable of receiving, for example, sensor data and / or data from one or more measuring devices of the production plant via a wired and / or wireless communication connection) and / or a mobile device, such as a tablet computer, a smartphone or the like, which is constructed to perform at least a part of the method and is connected to the processing device and / or the production plant via, for example, a wireless communication connection. As further disclosed in this specification, the method according to the aspect may be used in and / or for controlling at least part of a production plant for manufacturing at least one building material panel, or the method may at least support such control.

[0020] The production of building panels (sometimes referred to as "material panels" in the professional field) can be carried out either intermittently or continuously. In the intermittent production process, building panels are produced as flat objects with limited dimensions in all three spatial directions, while building panels produced in the continuous production process are cut-to-size pieces of web-like material with limited dimensions only in two spatial directions. The operating mode of the joining and / or compacting unit determines whether the entire process is described as an intermittent process or a continuous process, because in the so-called intermittent production method, the process steps upstream of the compaction, such as spreading, are usually also designed as continuous sub-processes. Since the compacting unit or the combined joining and compacting unit usually also works with great pressure when producing building panels, technicians usually refer to these units as "pressing parts" relative to the entire device. In the production of the building panels referred to in this article, the working pressure here is usually about 50N / cm2, depending on the material and size of the material board to be produced. 2 and about 500N / cm 2 Between, preferably 100N / cm 2 and 400N / cm 2 Although in the production of building material panels for buffering, so-called buffer panels, the working pressure can also be lower than 50N / cm2 at very low densities. 2 .

[0021] Building panels having at least one layer containing natural-based fiber components are of particular importance in building panels, both from an economic point of view and from a technical practicality point of view. Natural-based fibers or fiber components in the sense of this document are fibers and fiber components of natural origin, i.e. originating from annual or perennial plants, whether these fibers or fiber components are present as pure fibers, e.g. for the production of medium-density fiberboard / high-density fiberboard or veneer types thereof, or as components forming wood chips, long wood chips or wood flakes, which are usually used for the production of particleboard or oriented strand board (OSB) or veneer types thereof. Therefore, the term "wood particles" used hereinafter always includes at least natural fibers or fiber components.

[0022] Such building panels are often referred to by the person skilled in the art as "wood building panels" for short, even if they have one or more layers that are not based on raw materials obtained from perennial plants. Even building panels that contain only one or more layers that consist at least partly of fibers and / or fiber components obtained from annual plants are often referred to as wooden building panels and only in rare cases are they correctly referred to as "bast or grass building panels".

[0023] Such wooden building panels are implemented in a variety of forms and are used in a variety of applications. Particleboards, OSB boards, MDF (medium density fiberboard) or HDF (high density fiberboard) boards, as well as hybrid boards made from individual layers of such composite materials are particularly common. The name of the building panel depends on the shape and size of the fibers or particles used in the board structure or layer structure. If it is made from "fine" wood particles, the person skilled in the art will say it is a particleboard, while OSB boards are made from "coarse" wood particles. By "fine" wood particles, the person skilled in the art generally understands particles whose maximum extension in one spatial direction does not exceed 60 mm; in most cases, the maximum extension of these particles, which are called wood chips, is even a maximum of 25 mm or even 20 mm. By "coarse" wood particles, the person skilled in the art generally understands wood particles whose maximum extension in one spatial direction is at least 60 mm, and generally, the maximum extension of these wood particles, which are called long wood chips, is a maximum of 60 mm to 185 mm, in particular 80 mm to 140 mm.

[0024] On the other hand, medium density fiberboard and high density fiberboard or their individual layers are formed from (medium density or high density compressed) fibers which are usually obtained from raw materials by an intermediate introduction chemical process, usually a cooking process.

[0025] Hybrid panels consist of several layers of different types and are often particularly suitable when the material panel needs to meet different requirements for its use.

[0026] Such building panels are referred to as wood building panels even if they consist of a single layer without any natural-based fibers and / or fiber components. They are usually clad building panels, i.e. wood building panels with an external cladding on one or both sides. Plastic is usually used for the cladding. So-called coated particleboards are particularly known for this.

[0027] Thus, wooden building panels of the above-mentioned kinds and types are produced from wood particles (wood chips, long chips or fibers) of different shapes and sizes, which are bonded together under pressure and temperature in the so-called pressing section of the building panel production plant by activating the wood particles' own bonding mechanisms and adding an adhesive (usually glue).

[0028] Recently, in addition to wood materials that take many years to regrow, people have begun to pursue the use of annual plants, especially grasses, to produce building material panels. These annual plants have the great advantage of growing quickly. Therefore, the use of these plants is particularly resource-friendly and more in line with the growing environmental awareness around the world. In addition, the growing prosperity of many countries, such as Asian countries, requires a huge demand for building material panels to meet housing construction, especially in interior decoration and furniture manufacturing.

[0029] Since the annual plant has no bark, its harvested product initially forms a homogeneous raw material in terms of production technology, from which the fibers can be obtained for the production of building material panels by a fiber splitting process.

[0030] However, the processing of annual plants is much more complicated than the processing of building panels based on wood particles. During production, the large amount of silicates that have an abrasive effect on the plant structure is a major obstacle. This requires a considerable increase in the workload, especially in plant construction, for example through additional process steps, reinforcement of certain plant components and an increased need for spare parts. In addition, there is the risk of production downtime.

[0031] Furthermore, the mechanical properties of boards (layers) based on particles produced from annual plants are very different from those of boards (layers) based on wood particles.

[0032] The large number of process steps superimposed on one another, the different requirements placed on the individual process steps depending on the type of building panel to be produced and its required properties, as well as the parameters which influence the quality of the building panel to be produced in a process step, result in a concomitant structure of almost arbitrarily complex nature.

[0033] In an exemplary embodiment, the building material for producing the above-mentioned building material panel includes wood material, buffer material and / or one or more raw materials, such as straw, bagasse, bamboo, hemp, oil palm. Typical building material panels include particle board, coarse particle board or OSB board and / or medium density fiberboard.

[0034] For the purposes of the present disclosure, the term "building material panel" is understood to mean a board comprising, for example, at least a large proportion (e.g. more than 25% by weight) of at least one of the above-mentioned building materials, in the case of a wood material panel, having, for example, at least a large proportion (e.g. more than 50% by weight, in particular more than 75% by weight, for example at least 80% by weight) of cellulose-containing materials, for example lignocellulose-containing materials, such as sawdust. Such building material panels may also include other materials, such as plastic materials, which are, for example, at least partially present in the form of particles and / or fibers.

[0035] Thus, a production plant for manufacturing building material panels may in particular include a production plant for manufacturing wood material panels (or wood alternative building material panels). Such a production plant generally includes a plurality of production units or sections, such as a section for storage, debarking and / or chipping (e.g. a silo), a section for washing wood chips, a section for degreasing and / or chipping and gluing, a section for drying fibers and / or wood chips, a section for matting, and / or a section for hot pressing.

[0036] In particular, the production conditions of each section and / or unit of the production equipment can be characterized by corresponding process parameters. It should be understood here that in the sense of the present disclosure, the process parameters correspond in particular to actual values ​​and / or target values. Therefore, in an exemplary embodiment, the production equipment has at least one sensor and / or at least one measuring device. For example, the production equipment may have one or more pressure sensors, temperature sensors and / or humidity sensors. For example, the production equipment may further include one or more measuring devices for measuring the characteristics of the building material panel, such as for measuring the characteristics of wood chips. However, it is understood that the present disclosure is not limited to such sensors and / or measuring devices. In particular, the production equipment has at least one production section comprising at least one sensor. For example, the pressing section may have a pressure sensor, wherein the corresponding pressure measurement value may correspond to the process parameter pressure. Alternatively or additionally, the pressure setting target value of the corresponding pressing section may also correspond to the process parameter pressure.

[0037] In other words, in an exemplary embodiment, the method is performed by a control device of a production plant or by a system comprising the control device, wherein the production plant has at least one sensor and / or at least one measuring device, which is configured to output at least one sensor measurement value and / or at least one measuring device measurement value as a target value or actual value of a corresponding process parameter and characterize the corresponding process conditions during the production of building material panels by the production plant.

[0038] Exemplary process parameters of the present disclosure, especially process parameters of the fiberization and gluing sections, may include the amount of extrusion water, the amount of steam added to the digester, the filling level of the digester, the temperature of the digester, the steam pressure of the digester, the cooking time of the digester, the amount of sawdust, the amount of paraffin added, etc. Exemplary process parameters of the present disclosure, especially process parameters of the matting section, may include the amount of fiber discharge, the spreading height, the speed of the forming belt, the mat weight, the moisture content of the mat, the spreading width, the pre-pressing pressure, the pre-pressing distance, the trimming width, the mat density, the amount of water sprayed, etc. It goes without saying that other units and / or sections of the production equipment have their own process parameters, which characterize the corresponding production conditions of these other units and / or sections.

[0039] In a production facility, the production conditions that are present / set for / when producing building material panels can at least partially influence the properties of the produced building material panels. These properties of the building material panels can be characterized in particular by quality features or quality parameters. In the sense of the present disclosure, quality features or quality parameters can in particular include transverse tensile strength, bulk density, bending strength and / or thickness expansion, but these examples should not be understood as limiting. Further quality features or quality parameters can also be provided.

[0040] The production plant can have suitable sensors and / or measuring devices at different production stages in order to record the production conditions at different production stages during the production of building material panels and to assign them to the produced building material panels. In this way, sensor data with corresponding time stamps and / or measurement data from the measuring devices can be assigned to the produced building material panels and stored as data sets. These time-assigned data sets / data vectors contain the process conditions or process states when producing the corresponding building material panels.

[0041] After the production process of building material panels or their components is completed, the produced building material panels are usually ejected at regular time intervals in order to measure the corresponding quality characteristics or quality parameters of these ejected building material panels, for example in a laboratory or by a test robot. In this way, a sample data set can be obtained, which includes quality parameter data and corresponding process parameter data of building material panel samples (whole panels or partial panels) of the corresponding building material panel type. Therefore, these sample data sets provide an actual correlation between the set / existing process conditions of the corresponding section of the production equipment (which are represented by the process parameter data distributed in time) and the resulting quality characteristics of the building material panels produced under these process conditions. Based on the information contained therein, the sample data set can be used as a training data set for a mathematical model, which is used in the production of building material panels of the corresponding building material panel type, and is referred to as a training data set in the following.

[0042] Therefore, according to the aspect of the present invention, a training data set of a building material panel type includes quality parameter data representing at least one value of at least one corresponding quality characteristic of a building material panel sample of the building material panel type (for example, a measurement value measured in a laboratory for the corresponding building material panel sample), and process parameter data representing values ​​of multiple process parameters of a production method for producing the building material panel sample (or process parameter values, such as one or more target values ​​or actual values ​​for each process parameter, respectively, which are obtained, for example, by corresponding sensors, corresponding measuring devices and / or by corresponding setting devices of the production equipment).

[0043] In an exemplary embodiment, the process parameter data includes respective timestamps, or the process parameter data is assigned respective timestamps (e.g., stored in correspondence with the process parameter data), which timestamps correspond to respective process parameters. In other words, in an exemplary embodiment, a process parameter of the plurality of process parameters may correspond to a temporally assigned process parameter value including a timestamp. The timestamp corresponds to a point in time when the process parameter value characterizes a process condition existing during the production of the produced building material panel.

[0044] In an exemplary embodiment, the method comprises a step of providing (eg storing) at least one training data set on a storage medium which is connected to the production device or is configured for connection to the production device and which is accessible to at least one control device.

[0045] Based on the training data set, the manufacturing process of the building material panel type can be modeled for the production equipment. In an exemplary embodiment, for example, corresponding coefficients or weights of the mathematical model can be assigned to the process parameters, in particular for a plurality of training data sets, and these coefficients or weights can be determined in the corresponding equation system of the mathematical model based on the measured quality characteristics. This can be done continuously for more and more subsequent training data sets (for example, each time a new sample data set is obtained from laboratory measurements as a training data set during the production process), so that the mathematical model can be "learned" by more training data sets and thus "trained". Therefore, as mentioned above, the sample data set used to train the mathematical model (on which the mathematical model is based) is referred to as the training data set.

[0046] As described above, the method according to said aspect of the invention comprises preparing a mathematical model for use in the manufacturing process of at least one building material panel (e.g., a plurality of series of building material panels of a corresponding building material panel type), wherein the mathematical model is based on a number of training data sets, in particular training data sets of building material panel types for the manufactured building material panels and / or for building material panels that are part of the series of manufactured building material panels, the number corresponding to a minimum number or being greater than the minimum number.

[0047] In an exemplary embodiment, the mathematical model includes a set of model coefficients for one or more mathematical equations. In particular, in an exemplary embodiment, the mathematical model includes a set of model coefficients for one or more mathematical equations, wherein the mathematical equation or equations having the model coefficients are arranged such that the mathematical equation or equations can output a value (expected value) of at least one quality parameter based on input values ​​from at least one process parameter data set, the process parameter data set including process parameter values ​​with associated time stamps.

[0048] In other words, for example, target values ​​and / or measured values ​​of process parameters of the production plant set and / or measured during the production of the building material panels can be transmitted to at least one control device, which is connected to the production plant via a suitable communication connection. Using the process parameter values ​​thus obtained, the at least one control device can use a mathematical model to calculate predicted values ​​of quality characteristics that are expected for the building material panel just produced based on the set and / or measured target values ​​and / or actual values. Accordingly, the method according to the aspect comprises obtaining prediction data based on the mathematical model, wherein the prediction data represents a prediction of at least one quality parameter of at least one building material panel (produced building material panel and / or as part of a series of produced building material panels).

[0049] Here, in an exemplary embodiment, obtaining the prediction data comprises calculating the prediction data by and / or on at least one control device, for example by one or more processors and / or by one or more data processing systems. Alternatively or additionally, in an exemplary embodiment, obtaining the prediction data comprises receiving the prediction data by at least one control device, for example receiving the prediction data from a device connected to the control device at least via a communication connection.

[0050] These prediction values ​​can be displayed to an operator of the production plant so that the operator can influence the expected quality characteristics by readjusting the process parameters if necessary. Alternatively or additionally, control signals can also be generated based on the prediction data, based on which the at least one control device can correspondingly readjust the process parameters for producing the at least one building material panel.

[0051] Thus, in one exemplary embodiment, preparing a mathematical model for use in the production of at least one building material panel, wherein the mathematical model is based on a minimum number of training data sets, comprises storing on a storage medium connected to at least one production device and / or on a storage medium accessible to at least one production device at least one selected from the group consisting of:

[0052] - at least one model coefficient for one or more mathematical equations;

[0053] - at least one training data set for training and / or generating a mathematical model;

[0054] -Data representing mathematical formulas, such as program data.

[0055] In one exemplary embodiment, the training data set (the minimum number of training data sets) upon which the mathematical model is based includes a training data set used to train or generate a mathematical model for use in the manufacture of at least one building material panel.

[0056] Here, a mathematical model for use in the production of at least one building material panel can be generated based on the number of available training data sets for the type of building material panel to be produced (stored on a storage medium connected to at least one control device), wherein the mathematical model generated thereby can then be used to predict at least one quality parameter of the building material panel to be produced and / or being produced. In particular, if one or more building material panels are taken out of the ongoing production, measured in a laboratory, and thus a new sample data set is obtained as a training data set, the mathematical model can be updated based on the new training data set. For this purpose, the mathematical model can be generated based on the previously existing training data set and the new training data set, which can also be understood within the scope of the present disclosure as training the mathematical model based on the new training data set.

[0057] Thus, in one exemplary embodiment, the method comprises

[0058] - generating and / or training a mathematical model for use in the manufacture of at least one building material panel based on a number of training data sets that is equal to or greater than a minimum number.

[0059] In an exemplary embodiment, the mathematical model includes a statistical model or corresponds to a statistical model. In particular, in an exemplary embodiment, the mathematical model is based on a so-called interdependent synchronous stochastic linear equation model. It has been shown that, for example, a regression model of an interdependent synchronous (linear) stochastic equation system can be suitable for modeling the production process of building material panels. In particular, processes in the wood material industry can be regarded as interdependent and synchronous because multiple product characteristics are produced simultaneously, rather than independently of each other, and there are interdependencies between influencing variables and process parameters. The model equation can be described as linear because the model coefficients are linear. However, it will be understood that the present disclosure is not limited to such a linear system. For example, process parameters can also be entered with quadratic, cubic or exponential (or other) weights.

[0060] Therefore, in an exemplary embodiment, the mathematical model includes a simultaneous equation model, in particular, a mathematical model based on three-stage least squares ("Three-stage least squares", "3SLS"). In an exemplary embodiment, the mathematical model also includes a mathematical model based on two-stage least squares ("Two-stage least squares", 2SLS), a mathematical model based on partial least squares (PLS) and / or a linear regression model.

[0061] As previously mentioned, in an exemplary embodiment, the training data set for access by the at least one control device is stored (kept ready) on a storage unit that is accessible by the at least one control device. In an exemplary embodiment, the method according to the aspect includes, for example, maintaining / storing a plurality of training data sets on a storage medium to which the at least one control device is connected and / or to which the at least one control device is accessible. In an exemplary embodiment, the value of the process parameter represents a corresponding process condition of a production method for manufacturing a building material panel.

[0062] It is to be understood that data in the sense of the present disclosure, such as quality parameter data, process parameter data and / or data disclosed hereinafter, comprises any kind of information that is read and / or processed by a processing device, such as read and / or processed by at least one control device, a processor, a system of one or more processors and / or one or more computers, and in particular information stored for this purpose on a storage medium that is directly and / or indirectly (e.g. directly, by wire or wirelessly) connected in a suitable manner to at least one control device, a processor, a system of one or more processors and / or one or more computers.

[0063] As mentioned above, the mathematical model is based on a number of training data sets that is equal to or greater than a minimum number. Here, the minimum number may correspond to such a number of training data sets that is large enough so that the training / generation of the mathematical model with this number of training data sets will produce a mathematical model that can output a predicted value of at least one quality parameter, the quality of which meets a predetermined quality standard. In order to check whether the predicted value of the mathematical model meets the predetermined quality standard, the quality standard value predicted based on the process parameter value of the sample data set can be compared with the quality standard actual measured value of the sample data set. For example, if the absolute value of the difference between the predicted value and the actual measured value is less than or equal to the predetermined value, the quality standard can be met.

[0064] In other words, in an exemplary embodiment, the minimum number corresponds to such a number of training data sets that the mathematical model trained and / or generated based on this number of training data sets can output a predicted value of at least one quality criterion, wherein the quality of the predicted value meets the predetermined quality criterion. In an exemplary embodiment, for the quality parameter(s) for which the predicted data are obtained according to the mathematical model, the minimum number of training data sets is between 20 and 100, between 30 and 80, between 35 and 70, and / or between 40 and 60.

[0065] The minimum number can be fixed, for example for a mathematical model for a type of building material panel, or can be flexibly set, for example by user input. The latter case provides the operator with the option of flexibly setting the minimum number. In both cases, the minimum number value accessible to the at least one control device can be stored on a storage medium connected to the at least one control device and / or accessible to the at least one control device. In other words, in an exemplary embodiment, the method comprises:

[0066] - obtaining at least one information representing a minimum quantity value, wherein the obtaining comprises at least one of the following:

[0067] - acquiring (eg receiving) at least one information according to an input via a user interface connected to at least one device (at least one control unit);

[0068] - accessing a storage unit connected to at least one device (at least one control unit) to obtain at least one information;

[0069] According to said aspect of the invention, the number of training data sets on which the mathematical model is based comprises at least one training data set for (the) building material panel type of at least one building material panel (a produced building material panel and / or a building material panel that is part of a series of produced building material panels).

[0070] For frequently produced and / or mass-produced building material panel types, which account for a correspondingly high proportion of the total production, training / generating a mathematical model based on a training data set for only one building material panel type, for a building material panel type among the frequently produced and / or mass-produced building material panels, can in some cases achieve the minimum number relatively quickly. For such building material panel types, due to the large production volume, the possible number of building material panel samples that can be provided to the laboratory for measurement is also correspondingly large, so that the sample data set can be generated at a relatively high frequency, so that the minimum number can be achieved relatively quickly.

[0071] Therefore, the present disclosure includes an exemplary embodiment according to which the number of training data sets on which the mathematical model is based includes training data sets for only one (the) building material panel type in at least one building material panel (the building material panel has been manufactured and / or is part of a series of building material panels that have been manufactured).

[0072] It is understood that, as long as the minimum quantity has not been reached for a certain building material panel type and the mathematical model of the building material panel type cannot be used for the production of the corresponding building material panels to predict the expected quality parameters, the production of the building material panel type can be carried out without such prediction values ​​based on the empirical values ​​of the production equipment operators. However, it has been found that controlling the production equipment based on the prediction values ​​of the mathematical model (e.g., by the equipment operator directly controlling the equipment based on the displayed prediction values ​​and / or through control data generated based on the prediction data and used to directly / automatically / electronically control at least a part of the production equipment) can improve the production efficiency of the building material panels.

[0073] Therefore, it is desirable to use the mathematical model also when producing the following types of building material panels, that is, such building material panels have not yet reached the minimum number of training data sets, for example due to a low production proportion in the total output of the production facility. Here, it was found that the mathematical model can be generated / trained based on training data sets, which are only partially used for the relevant building material panel type and partially used for other building material panel types. In other words, in an exemplary embodiment, the number of training data sets includes at least one training data set for the at least one building material panel type and at least one training data set for at least one other building material panel type.

[0074] It has been found that it is particularly feasible if the building material panel types with a lower production ratio are grouped with the building material panel types with a higher production ratio, i.e. the number of training data sets consists of the number of training data sets of the building material panel types with a lower production ratio and the number of training data sets of the building material panel types with a higher production ratio. However, grouping is also possible and generally reasonable if the production ratios are approximately the same, for example differing by no more than 10%. Therefore, in an exemplary embodiment, the production ratio of building material panels of the building material panel type of at least one building material panel to the total production volume is approximately equal to or less than the production ratio of building material panels of the other building material panel types. In other words, in an exemplary embodiment, the mathematical model is based on several training data sets of the building material panel type of at least one building material panel (produced and / or part of a series of building material panels produced) and at least on several training data sets of other building material panel types, wherein the number of training data sets of the building material panel type of at least one building material panel (produced and / or part of a series of building material panels produced) is less than the minimum number and / or less than the number of building material panels of the other building material panel types. Here, the total number of training data sets is equal to or greater than a minimum number, i.e., the number of training data sets of a building material panel type of at least one building material panel (produced and / or part of a series of building material panels produced) plus the number of training data sets of other building material panel types is equal to or greater than the minimum number.

[0075] It has been found that for panel types for which there is no sufficient training data set available (whose number is less than a minimum number), it may be advantageous to group them with panel types for which there is sufficient training data set available (whose number is equal to or greater than the minimum number), especially when the panel types are similar in at least one characteristic. Panel types may be characterized and / or classified in particular according to their characteristics, wherein, in an exemplary embodiment, the characteristics of the building material panel types are at least selected from the following aspects:

[0076] - Thickness of building material panels;

[0077] - Width of building material panels;

[0078] - Building material panel density;

[0079] - Type of gluing of building panels;

[0080] - at least one material of the building material panel, in particular at least one glue and / or one wood material.

[0081] The building material panel may be characterized by one or more of these characteristics, and in an exemplary embodiment, corresponding parameter values ​​of one or more of these characteristics are accessible to at least one control device, such as stored on a storage medium connected to and / or accessible to at least one control device.

[0082] In an exemplary embodiment, the building material panel types are similar in at least one characteristic if at least one parameter value representing the at least one characteristic is similar for different building material panel types. This is the case if the difference between the parameter values ​​of the two building material panels in terms of the at least one characteristic does not exceed a maximum value, such as a predetermined maximum value, such as a maximum value defined by an operator. Thus, for at least one building material panel type (produced and / or part of a series of building material panels produced) and at least one other building material panel type, the difference between at least one parameter value characterizing the characteristic of the building material panel type of the at least one building material panel (produced and / or part of a series of building material panels produced) and at least one corresponding parameter value characterizing the response characteristic of at least one other building material panel type is equal to or less than the maximum value. Here, in an exemplary embodiment, the value comprises an absolute value or an amount / absolute amount, but may also be other values ​​that take into account the distance between two parameters.

[0083] As mentioned above, it is advantageous to use a mathematical model based on one of the types of building material panels to be produced, so it is also advantageous to set one building material panel type for at least one control device, for example when starting to produce a series of building material panels of the same type. Therefore, in an exemplary embodiment, the method includes:

[0084] - Acquiring (eg receiving) information representing a building material panel type of at least one building material panel (produced and / or part of a series of produced building material panels).

[0085] It is understood that information within the meaning of the present disclosure, such as information representing the type of building material panel of at least one building material panel and / or information disclosed below, includes any type of computer-readable and / or electronically readable information that can be read and / or processed by a processing device, such as at least one control device, a processor, a system consisting of one or more processors and / or one or more computers.

[0086] In an exemplary embodiment, information representing a building material panel type of at least one building material panel (manufactured and / or part of a manufactured series of building material panels) is obtained based on input made through a user interface. In an exemplary embodiment, the user interface includes:

[0087] - at least one keyboard and at least one screen,

[0088] - one or more touch screens,

[0089] - one or more voice input devices, and / or

[0090] -Mobile devices connected to production equipment, such as smartphones, tablets, and laptops.

[0091] In an exemplary embodiment, the user interface is directly connected to the production equipment and / or the at least one control device by wired and / or wireless means and / or, for example, via the Internet. In an exemplary embodiment, the user interface comprises a display device such as one or more screens and an input device such as one or more keyboards, a computer mouse and / or an input / output device such as one or more touch screens. In an exemplary embodiment, the at least one control device comprises a mobile device such as one or more smartphones, one or more tablets and / or one or more laptops, wherein the user interface comprises, in particular in these embodiments, one or more touch screens of the mobile device.

[0092] Based on the data representing the building material panel type of at least one building material panel, at least one control device can check whether a mathematical model exists for the building material panel type to be produced (for example, stored on a memory connected to the at least one control device), which mathematical model is based on a sufficient training data set (only for the building material panel type to be produced or grouped with another building material panel type).

[0093] Accordingly, in one exemplary embodiment, the method comprises:

[0094] - determining, based on the obtained information representing the building material panel type of the at least one building material panel, whether there is a mathematical model that can be used in the production of the at least one building material panel, the mathematical model being based on a number of training data sets that is equal to or greater than a minimum number.

[0095] Here, it is understood that determining whether a mathematical model exists includes determining whether such a mathematical model (based on a sufficient training data set) is stored, for example, on a storage unit connected to at least one control device and / or accessible to at least one control device. It is further understood that the connection according to the present disclosure particularly includes a direct or indirect wired communication connection (e.g. a local area network connection), and / or a direct or indirect wireless communication connection, including a radio connection, such as Bluetooth, NFC, WLAN, 4G or 5G and / or a communication connection via the Internet. An indirect connection may include one or more connections via one or more intermediate nodes.

[0096] It can be further understood that, in an exemplary embodiment, storing the mathematical model includes storing at least one selected from the following:

[0097] - at least one model coefficient of one or more mathematical equations for the mathematical model;

[0098] - at least one training data set for training and / or generating a mathematical model;

[0099] -Data representing mathematical equations, such as program data.

[0100] If, based on the obtained information representing at least one type of building material panel, it is determined that there is a mathematical model that can be used in the production of at least one building material panel, the mathematical model being based on a number of training data sets that is equal to or greater than a minimum number, then in one exemplary embodiment, the method further comprises:

[0101] When it is determined that a mathematical model exists that can be used in the production of the at least one building material panel, the mathematical model is prepared for use in the production of the at least one building material panel.

[0102] Otherwise, in one exemplary embodiment, no mathematical model is used in manufacturing the at least one building material panel.

[0103] In other words, in this exemplary embodiment, the step of preparing the mathematical model for use in the production of the at least one building material panel includes preparing the mathematical model for use in the production of the at least one building material panel upon determining that the mathematical model can be used to manufacture the at least one building material panel.

[0104] As described above, in some cases, the mathematical model provided for a type of building material panel may output predicted values ​​that degrade in quality over time. For example, such changes may be the result of changing seasons, changes in production equipment components and / or materials, etc. However, it has been advantageously discovered that the effects of such long-term changes can be effectively mitigated by no longer using old training data sets to generate / train mathematical models used in the production of at least one type of building material panel.

[0105] In particular for this purpose, in an exemplary embodiment, the training data set comprises a timestamp, and / or in this exemplary embodiment, there is a corresponding association between the timestamp and the training data set. The timestamp represents the point in time when the building material panel sample was produced. For example, this point in time may be the point in time when the building material panel sample was sawed off from a longer building material panel strip in production, and / or the point in time when the building material panel sample was ejected for measurement in the laboratory. It goes without saying that the timestamp represents such a point in time, but the specific value is not important. For example, there may be a particularly fixed time difference between the actual point in time when the building material panel sample was ejected and / or sawed off and the time value represented by the timestamp. Such a timestamp can cause the training data set to be discarded when it is no longer the latest data set.

[0106] In an exemplary embodiment, the method further comprises:

[0107] - obtaining information representing that at least one training data set on which the mathematical model is based comprises or corresponds to a timestamp representing a point in time before a predetermined time interval.

[0108] The information may be automatically obtained from at least one control device and / or based on user input. Therefore, in an exemplary embodiment, obtaining information indicating that at least one training data set based on the mathematical model includes a timestamp representing a time point before a predetermined time interval or a statement corresponding to the timestamp includes at least one of the following:

[0109] - according to an input through the user interface, obtaining information indicating that at least one training data set on which the mathematical model is based includes a timestamp representing a time point before a predetermined time interval or a statement corresponding to the timestamp;

[0110] - determining whether at least one training data set on which the mathematical model is based comprises or corresponds to a time stamp representing a point in time before a predetermined time interval.

[0111] Therefore, in one exemplary embodiment, the method further comprises:

[0112] - if at least one training data set comprises and / or corresponds to a time stamp representing a point in time before a predetermined time interval, discarding the at least one training data set from the number of training data sets on which the mathematical model is based.

[0113] In an exemplary embodiment, the predetermined time interval comprises a time interval before starting to perform the method and / or starting to produce the at least one building material panel. In an exemplary embodiment, the time interval is determined in relation to starting to perform the method and / or starting to produce the at least one building material panel. For example, the time interval may be a number of years (e.g., 2 years) before the date when the production of the at least one building material panel starts using the mathematical model. In an exemplary embodiment, the predetermined time interval thus comprises one or more days, one or more months, and / or one or more years before the production of the at least one building material panel starts using the mathematical model.

[0114] In an exemplary embodiment, the method further comprises:

[0115] - generating or training a mathematical model for use in the production of at least one building material panel based on the number of training data sets excluding the discarded training data sets.

[0116] In one exemplary embodiment, preparing a mathematical model for use in the production of at least one building material panel includes preparing a mathematical model for use in the production of at least one building material panel, wherein the mathematical model is based on a number of training data sets that do not include discarded training data sets.

[0117] In this way, it is advantageously achieved that old training data sets are no longer used for training and / or generating the mathematical model, so that the mathematical model is kept up to date and thus long-term negative influences, such as seasonal influences, are reduced. In this way, the mathematical model can in particular output more reliable forecast data.

[0118] As mentioned above, the mathematical model used in the production of at least one building material panel can be generated based on the number of training data sets available (stored on a storage medium connected to the at least one control device) for the type of building material panel to be produced. In particular, if one or more building material panels are ejected from the ongoing production and measured in the laboratory, and thus a new sample data set is obtained as a training data set, the mathematical model can be updated based on the new training data set. Similarly, the mathematical model can also be updated, for example, before the production of at least one building material panel starts, for example before the production of a series of building material panels of one building material panel type starts.

[0119] For this update process, in an exemplary embodiment, the method includes:

[0120] - Acquiring information representative of the statement that there is at least one training data set (eg stored on a storage medium connected to and / or accessible to at least one control device) which is not included in the number of training data sets on which the mathematical model is based.

[0121] In other words, in this embodiment, the at least one control device is configured to obtain information about the existence of other training data sets in addition to the training data set on which the mathematical model is based. In an exemplary embodiment, the step of obtaining information about the existence of at least one training data set that is not included in the number of training data sets on which the mathematical model is based is performed before and / or during the production of at least one building material panel using the mathematical model.

[0122] When information is obtained that there is at least one training data set that is not included in the number of training data sets on which the mathematical model is based, in an exemplary embodiment, the method includes:

[0123] - training and / or generating a mathematical model based on the training data sets included in the number of training data sets and the at least one training data set not included in the number of training data sets.

[0124] Therefore, in this case, the mathematical model can be generated based on the previously existing training data set and the new training data set that has not yet been included in the mathematical model. This can be done before starting to manufacture at least one building material panel and / or during the manufacturing of at least one building material panel (during the manufacturing process). In the present disclosure, generating a mathematical model based on the previously existing training data set and the newly added training data set can also be understood as training the mathematical model based on the new training data set.

[0125] The information that there is at least one training data set that is not included in the number of training data sets on which the mathematical model is based can be provided by an operator of the production facility via an input in the user interface and / or automatically detected by at least one control device. Thus, in one exemplary embodiment, the method comprises:

[0126] - Based on an input via the user interface, obtaining information indicating the presence of at least one training data set that is not included in the number of training data sets on which the mathematical model is based.

[0127] Alternatively or additionally, obtaining information indicating the presence of at least one training data set that is not included in the number of training data sets on which the mathematical model is based comprises:

[0128] - determining whether there is at least one training data set that is not included in the number of training data sets on which the mathematical model is based.

[0129] In this embodiment, training and / or generating the mathematical model according to the training data set included in the number of training data sets and according to the at least one training data set not included in the number of training data sets includes:

[0130] - if it is determined that there is at least one training data set which is not contained in the number of training data sets on which the mathematical model is based, the mathematical model is trained and / or generated based on the training data sets contained in the number of training data sets and based on the at least one training data set which is not contained in the number of training data sets.

[0131] In this exemplary embodiment, the method further includes obtaining (e.g., calculated by at least one control device) prediction data based on a mathematical model, wherein the mathematical model is based on a training data set included in the number of training data sets and on at least one training data set not included in the number of training data sets, wherein the prediction data represents a prediction of at least one quality parameter of at least one building material panel (a manufactured building material panel and / or a part of a manufactured building material panel series), and the method also includes outputting the prediction data.

[0132] The updating process based on new training data sets enables the model to be kept up to date, in particular during the production of building material panels of a certain building material panel type, so that the model can be increasingly adapted to this building material panel type. This is particularly advantageous if the mathematical model is initially based on a grouping of the training data set for the building material panel type to be produced with training data sets for other building material panel types. In this case, by continuously updating the mathematical model based on new training data sets obtained in the production process, it is possible that after a certain production time (possibly after several production cycles) there are sufficient training data sets for the building material panel type to be produced, so that a mathematical model based only on the training data set for the building material panel type to be produced can be used to predict this building material panel type.

[0133] A production facility for manufacturing building material panels usually includes a plurality of production sections and / or production units. During the production of building material panels, the production conditions to which the intermediate products are subjected in these production sections and / or production units may be determined by a plurality of process parameters. For example, since the corresponding sensors and / or measuring devices of the production facility determine the target values ​​and actual values ​​of each corresponding process parameter during the production process, and these target values ​​and actual values ​​are stored in the sample data set at least for each corresponding building material panel sample (whose quality characteristics are measured in the laboratory), the sample data set may contain target values ​​and actual values ​​of a large number of process parameters (for example, up to or even more than 3,500 process parameters in a typical production facility).

[0134] However, since not all process parameters need to have an equally large influence on the quality characteristics of the produced building material panels, it has proven to be advantageous to preselect (usually several hundred, for example 250 to 500) process parameters (to determine a set of process parameters) which are taken into account when adapting and / or generating the mathematical model (for example by at least one control device). In an exemplary embodiment, this preselection is provided to the at least one control device in the form of process parameter selection information. This can in particular control and, if necessary, minimize the amount of calculations required for adapting and / or generating the mathematical model. It has proven to be advantageous if the mathematical model outputs a predicted value for the preselected quality parameter.

[0135] Thus, in one exemplary embodiment, a mathematical model to be used in the manufacture of a building material panel includes:

[0136] - obtaining process parameter selection information representative of a selection of process parameters of a production method for manufacturing at least one building material panel, and / or obtaining information representative of at least one selected quality parameter, for which the value of the selected quality parameter is to be predicted by the mathematical model, for the at least one building material panel (the manufactured building material panel and / or a part of a manufactured building material panel series);

[0137] - Selecting a mathematical model according to process parameters and / or generating a mathematical model for at least one selected quality parameter.

[0138] Here, the process parameter selection information and / or the information representing at least one selected quality parameter is obtained by the at least one control device by querying a database, for example, via an Internet connection, and / or based on locally stored information, for example, information stored on a storage unit connected to the at least one control device. Alternatively or additionally, this information can also be obtained by user input using a user interface. In other words, in an exemplary embodiment, obtaining the process parameter selection information and / or the information representing at least one selected quality parameter includes at least one of the following:

[0139] - obtaining process parameter selection information and / or information representative of at least one selected quality parameter based on input made through a user interface;

[0140] - Retrieving process parameter selection information and / or information representative of at least one selected quality parameter based on a query of a database and / or at least one storage unit accessible to the control device.

[0141] In an exemplary embodiment, the method is performed by at least one control device of a production plant for manufacturing building material panels or a system comprising at least one such control device. Here, the production plant comprises at least one sensor and / or measuring device, which is configured to output at least one sensor measurement value and / or measuring device measurement value as a target value or actual value of a corresponding process parameter, which process parameter characterizes a corresponding process condition of the production plant during the production of the building material panel, wherein obtaining prediction data according to the mathematical model further comprises:

[0142] - Determining prediction data as a function of the mathematical model and at least one sensor measurement value and / or at least one measurement value of the measuring device.

[0143] In an exemplary embodiment, obtaining the prediction data comprises using at least one obtained sensor measurement value and / or at least one obtained measuring device measurement value as an input variable of the mathematical model, and calculating the prediction data by the mathematical model based on the input variables. Here, the sensor measurement value and / or the measurement value of the measuring device can be received by at least one control device, for example, via one or more corresponding communication connections with at least one corresponding sensor and / or at least one corresponding measuring device of the production plant. Therefore, in an exemplary embodiment, the method comprises:

[0144] At least one sensor measurement value and / or at least one measurement value of a measuring device is received via a communication connection from at least one sensor of a production plant for producing at least one building material panel and / or from at least one measuring device of a production plant for producing at least one building material panel.

[0145] As described above, the method according to the aspect includes outputting prediction data. As described above, the prediction data can be used to control a production device for producing at least one building material panel and / or at least support the control of the production device. In an exemplary embodiment, the method also includes:

[0146] According to the output prediction data, at least one unit and / or one section of a production equipment for producing at least one building material panel is controlled. It should be understood that in an exemplary embodiment, the control includes at least one of the following:

[0147] - directly controlling at least one unit and / or at least one section according to a corresponding control signal generated by at least one control device based on the output prediction data;

[0148] - the plant operator controls at least one unit and / or at least one plant section according to a display (eg a visual display on a display device) of the prediction of at least one quality parameter based on the outputted prediction data.

[0149] Thus, the output of the prediction data may comprise an internal output of the data, e.g. within the at least one control device, e.g. for further processing of the prediction data by the at least one control device. The output of the prediction data may further comprise outputting the prediction data for further processing by an external device (of the at least one control device). For example, the output prediction data may be converted in the control device or further data processing device into so-called control signals and / or display data, which display data may be used by said display device (e.g. a screen connected to the control device) for displaying a representation or diagram.

[0150] Thus, in one exemplary embodiment, the method may include:

[0151] - based on the outputted prediction data, causing display of a representation of a prediction of at least one quality parameter of at least one building material panel (the manufactured building material panel and / or as part of a series of manufactured building material panels) based on the prediction data; and / or

[0152] - generating a control signal based on the output prediction data for controlling at least one unit and / or one section of the production equipment, and / or for adjusting at least one process parameter of the production equipment during the production of at least one building material panel.

[0153] The representation here includes, for example, a graphical display of the predicted data in the form of a data value-time curve on a screen (e.g., a real-time display), wherein the screen can be directly or indirectly connected to the control device and / or the production equipment as a display device. Such display of the predicted data enables the user to adjust the production equipment according to the displayed predicted value of the quality characteristic (e.g., real-time adjustment).

[0154] In an alternative or additional embodiment, at least one control signal generated based on the prediction data is used to directly control at least one component of the production plant, for example using control electronics operating based on a feedback loop. BRIEF DESCRIPTION OF THE DRAWINGS

[0155] More advantageous exemplary embodiments of the present invention can be found in the detailed description of some exemplary embodiments of the present invention below, in particular in conjunction with the accompanying drawings. However, the drawings of the present application are only for illustrative purposes and do not limit the scope of protection of the present invention. The drawings are not necessarily drawn accurately to scale, and their purpose is only to reflect the general concept of the present invention by way of example. In particular, the features contained in the drawings should never be regarded as essential components of the present invention.

[0156] In the figure:

[0157] Figure 1 A schematic diagram of an exemplary production apparatus and control device for manufacturing building material panels is shown;

[0158] Figure 2 An exemplary flow chart is shown, which illustrates a method according to an exemplary embodiment of the aspects of the present invention;

[0159] Figure 3 A schematic diagram showing the grouping of building material panel types;

[0160] Figure 4 An exemplary flow chart is shown, which illustrates the steps of a method according to an exemplary embodiment of the aspects of the present invention; and

[0161] Figure 5 A schematic diagram is shown of an exemplary embodiment of a device, such as a mobile device, according to aspects of the present invention. DETAILED DESCRIPTION

[0162] Figure 1 A schematic diagram of an exemplary production apparatus 1 for producing a material panel made of wood chip material as an exemplary example of a building material panel according to the present disclosure is shown. Figure 1 Also shown is a schematic diagram of a control device 200, which is connected to the production equipment 1, for example, via the connection 550 shown in the schematic diagram, for controlling the production equipment 1. To this end, the control device 200 can be configured to perform the steps of the method according to the aspect of the present invention. To this end, the control device 200 can, for example, include a processing device, such as a computer and / or a computer system connected to sensors and / or measuring devices (not shown) of the production equipment 1, for example, to obtain target values ​​and / or actual values ​​of process parameters of each section or unit of the production equipment 1 from these sensors and / or measuring devices. Alternatively or additionally, in an exemplary embodiment, the control device 200 can include one or more mobile devices, such as one or more smartphones, one or more tablets and / or one or more laptops.

[0163] In an exemplary embodiment, control device 200 includes one or more display devices, such as one or more screens, and / or is connected to one or more display devices to display, for example, at least one quality characteristic of the building material panel being produced.

[0164] In an exemplary embodiment, the control device 200 may include multiple control devices, wherein a single control device may perform one or more steps of the method according to the aspects of the present invention. Figure 1The control device 200, which is only schematically shown in the figure, may include a processing unit, such as one or more computers connected (wirelessly, wired and / or via the Internet) to sensors and / or measuring devices of the production device 1, and a mobile device, such as at least connected (wirelessly, wired and / or via the Internet) to the processing unit. For example, in an exemplary embodiment, in particular the step of displaying and / or inputting data and / or information may include using a mobile device such as a smartphone (e.g. a touch screen) to display and / or input data.

[0165] Figure 1 Further shown is a storage device (e.g. one or more hard drives and / or one or more cloud storage units), which may, for example, be configured to store sample data sets and / or training data sets and for this purpose is connected to the control device 200 via a schematically illustrated connection 500.

[0166] In an exemplary embodiment, the connection according to the present disclosure, in particular the schematically illustrated connection 550 between the control device 200 and the production equipment 1 (for example, between the control device 200 and one or more sensors, measuring devices and / or control controllers of the production equipment 1) and 500, is in particular a direct or indirect wired communication connection (for example, a local area network connection), and / or a direct or indirect wireless communication connection, including a radio connection, such as Bluetooth, NFC, WLAN, 4G or 5G and / or a communication connection via the Internet.

[0167] Figure 1 The schematic diagram of the units, sections and devices of a production plant 1 for producing building material panels is shown. In particular, the following sections of the production plant 1 are shown, which can be used in the individual steps of the production of building material panels: a chipper 2, a chip dryer 3, a sorting device 4, a device for applying a binder to the chips with mixers 5a, 5b, which also represent other devices capable of applying a binder to the chips, and a spreading device 6, which spreads the glued chips, if necessary in the form of different layers of different chip sizes, onto a forming belt 7 by means of a plurality of spreading heads 6a, 6b, to form a mat, and furthermore, a continuous press 8 and a cutting or milling device 9 for adjusting the size of the finished building material panel.

[0168] The chipper 2 symbolically represents the entire process of cutting wood into wood chips. The diagram shows five ring-knife chippers, which are fed from the upper silo via a conveying screw. These chippers can produce wood chips of different sizes depending on the knife settings. The generated wood chips of different sizes and moisture contents are further sent to the dryer 3.

[0169] The subsequent sorting device 4 in the figure can be used in different embodiments at different production stages. The device can be used to filter out fine particles (such as dust) or large particles (such as unwanted minerals or adhesive agglomerates) from the wood chip flow. In addition, the sorting device 4 can also be used, for example, to sort the wood chips according to their size during the spreading process, so that different layers of the subsequent building material panel can be produced from wood chips of different sizes.

[0170] exist Figure 1 In the embodiment shown, for example two different chip size ranges are produced and conveyed from a silo to a mixer 5a and a mixer 5b, respectively. In these mixers, the chips are at least partially moistened with a binder.

[0171] Thus, a spreading head 6a for the top layer of the mat to be spread and pressed can be fed with chips of a different size than a spreading head 6b for the middle layer.

[0172] The basic structure of a double belt press 8 for manufacturing building material panels, in particular for producing building material panels of wood materials, comprises an upper press section with an upper heatable press plate and a lower press section with a lower heatable press plate. The upper and lower press sections are connected by a frame, and a pressure device for applying pressure is supported in the frame. In the upper and lower press sections, a continuously circulating steel belt is guided around a belt guide roller and forms a pressing gap to apply pressure and temperature to the mat.

[0173] In addition, Figure 1 A chip size measuring device 10 (an example of a measuring device) can also be seen, which is located after the mixer 5. Optionally, a chip size measuring device 11 can also be provided before the mixer. In these chip size measuring devices 10 and 11, selected representative chips can be measured, for example, a small sample of chips separated at a sampling point 20 during transportation. Exemplary process parameters related to the chips measured by these chip size measuring devices include: the amount of adhesive supplied through the adhesive nozzle, the amount of chips delivered, and / or the speed of the chips passing through the mixer 5a, 5b, the shaft speed of the mixer 5a, 5b, etc.

[0174] The method according to the above aspect of the present invention can be used in Figure 1 The production equipment shown is performed in the context of the present invention, but it should be understood that the present invention is not limited to Figure 1 The production equipment shown.

[0175] According to an exemplary embodiment, the process parameters, in particular the process parameters of wood material production, such as the process parameters of the fiberization and gluing section in a production plant for producing wood material boards, may include at least one or more of the following process parameters:

[0176] -Amount of squeeze water;

[0177] - amount of steam added to the digester;

[0178] - Digester liquid level;

[0179] - digester temperature;

[0180] - Digester steam pressure;

[0181] - cooking time;

[0182] - Amount of wood chips input;

[0183] - Paraffin wax addition amount;

[0184] -Refiner energy consumption;

[0185] -fiber mill temperature;

[0186] - Fiber mill steam pressure;

[0187] -Fiber mill grinding gap;

[0188] - Grinding disc life;

[0189] - Exhaust valve opening;

[0190] - pH value of fiber;

[0191] -Amount of glue used.

[0192] In an exemplary embodiment, the process parameters of the mat forming section in the production equipment for producing wood material boards may include at least one or more process parameters selected from the following:

[0193] - Fiber output;

[0194] - Spread height;

[0195] - forming belt speed;

[0196] - Basis weight of the pad;

[0197] - moisture content of the pad;

[0198] - Spread width;

[0199] - preload pressure;

[0200] -Preload spacing;

[0201] - Trimming width;

[0202] - pad density;

[0203] -Spray water volume;

[0204] - pad height at the end of the forming belt;

[0205] - pad temperature;

[0206] -Spraying height;

[0207] - Spreading unevenness.

[0208] In the production plant, corresponding sensors and / or measuring devices can be provided for such process parameters, so that correspondingly set target values ​​or actual values ​​can be provided for these process parameters via a corresponding communication connection of at least one control device.

[0209] In particular, for example, measured and / or set target values ​​or actual values ​​can be stored together with corresponding time stamps when producing building material panels, so that corresponding process parameter values ​​at various time points can be provided for the produced building material panels. As mentioned above, for those building material panel samples that are sent to a laboratory for measurement, for example, additional measurement values ​​for describing quality characteristics can be obtained and stored together with the process parameter values ​​in a sample data set or a training data set for the corresponding building material panel samples.

[0210] According to an exemplary embodiment, the quality characteristics that can characterize the building material panel may be one or more quality characteristics selected from at least the following:

[0211] - Transverse tensile strength;

[0212] - Bulk density;

[0213] - flexural strength;

[0214] -Thickness expansion ratio.

[0215] Thus, in an exemplary embodiment, a sample data set or training data set of building material panel samples includes data (process parameter data) representing at least one value (e.g., target value and / or actual value) of at least one corresponding process parameter with a corresponding time stamp, and data (quality parameter data) representing at least one value (e.g., measured value and / or laboratory measured value) of at least one corresponding quality characteristic of the building panel samples.

[0216] also, Figure 1 Furthermore, a storage medium 250 is shown, which is connected to the control device 200 via a connection 500 so that the control device can access the storage medium.

[0217] Figure 2 1 shows a flow chart representing an exemplary embodiment of a method 100 according to an aspect of the present invention. The flow chart 100 can be seen as a method for controlling a production device for producing building material panels, such as Figure 1The following description assumes that the method 100 is composed of the following exemplary control flow of the production equipment 1 shown in FIG. Figure 1 However, in other exemplary embodiments, the method 100 may be executed by one or more processors of the control device 200, and / or by multiple control devices, where, for example, one or more processors and / or one or more control devices may perform one or more steps of the method 100.

[0218] like Figure 2 As shown, method 100 includes step 101 of preparing a mathematical model for use in the production of at least one building material panel, wherein the mathematical model is based on a number of training data sets that equals or exceeds a minimum number. In other words, as described herein, method 100 may include, for example, a step of storing the mathematical model (e.g., storing model coefficients for one or more mathematical equations, a training data set for training and / or generating the mathematical model, and / or data, such as program data representing the mathematical equation).

[0219] As shown, the training data set for a certain building panel type includes quality parameter data, which represents at least one value of at least one corresponding quality characteristic of a building panel sample of the building panel type, and process parameter data, which represents values ​​of multiple process parameters in a production method used to produce the building panel sample.

[0220] As further shown, the number of training data sets includes at least one training data set for a building material panel type for the at least one building material panel. As described herein, in an exemplary embodiment, the number of training data sets may only include training data sets for a building material panel type for the at least one building material panel (i.e., a produced building material panel and / or a portion of a produced series of building material panels). In an exemplary embodiment, the number of training data sets may include at least one training data set for a building material panel type for the at least one building material panel (i.e., a produced building material panel and / or a portion of a produced series of building material panels), and at least one training data set for at least one other building material panel type. In an exemplary embodiment, the mathematical model may be based on the number of training data sets for the building material panel type for the at least one building material panel, and at least based on the number of training data sets for the other building material panel types, wherein the number of training data sets for the building material panel type for the at least one building material panel is less than a minimum number and / or less than the number of building material panels for the other building material panel types.

[0221] This is Figure 3 By way of example, four different board types A, B, C and D are shown. Figure 3Block 310 in the diagram illustrates a situation where building material panel type A corresponds to the at least one building material panel (i.e., a produced building material panel and / or a part of a produced building material panel series). As shown, there is sufficient training data set for panel type A, so a mathematical model can be generated based only on the training data set of building material panel type A for use in the production of building material panels of building material panel type A.

[0222] Block 330 illustrates a case where the number of training data sets for the building material panel type (building material panel type D in the illustrated case) of the at least one building material panel (i.e., the produced building material panel and / or part of the produced building material panel series) is less than the minimum number and less than the number of building material panels of the other building material panel types (building material panel type D in the illustrated case). However, since in this exemplary case, the building material panels of building material panel types A and D have the same characteristics in terms of sheet thickness and gluing method, in this case, grouping building material panel types A and D enables the generation of a mathematical model based on the training data set of building material panel type A and the training data set of building material panel type D, and is used to predict the quality characteristics of building material panel type D, even if sufficient training data sets have not been generated for the panel type alone. By using the described update process, the mathematical model based on the grouping of training data sets can be gradually updated with the training data set of building material panel type D for building material panel type D, until eventually no grouping is required for building material panel type D.

[0223] For building material panel type D, for which only a small amount of training data is available, the grouping shown of building material panel type A and building material panel type D is very advantageous because sufficient training data is available for building material panel type A alone. Figure 3 Box 320 in FIG. 3 shows the grouping of building material panel types B and C, both of which do not have sufficient training data sets individually in this example. However, these building material panel types B and C can still be grouped together because the sum of the number of training data sets that exist for the individual building material panel types B and C is greater than the minimum number.

[0224] Back to Figure 2 The method 100 further comprises the step of obtaining prediction data based on the mathematical model, wherein the prediction data represents a prediction of at least one quality parameter of the at least one building material panel (i.e., the produced building material panel and / or a part of the produced building material panel series). As further described and as Figure 2 As further shown, the method further includes step 103, namely outputting the prediction data.

[0225] Figure 4An exemplary flow chart 400 is shown, which represents the steps of an exemplary embodiment of a method according to the aspects of the present invention. The flow chart 400 can be understood as illustrating an exemplary control process for controlling a production equipment for manufacturing building material panels, such as Figure 1 The production equipment 1 shown. Without limiting the present invention, it is assumed that the method 400 is Figure 1 However, in other exemplary embodiments, the method 400 may be performed by one or more processors of the control device 200, and / or by multiple control devices, wherein, for example, one or more processors and / or one or more control devices may perform one or more steps of the method 400.

[0226] As shown, method 400 includes step 401, i.e., obtaining information representing the type of building panel of at least one building material panel to be manufactured. This step can be performed, for example, when the type of building material panel to be produced changes during production change. The information about the type of building material panel can be based on content input by a production equipment operator through a user interface, for example.

[0227] As further shown, the control device 200 will determine whether there is a mathematical model for the building material panel type based on this, and whether the number of training data sets based on the mathematical model is equal to or exceeds the minimum number. In other words, the control device 200 checks in step 402 whether there is a mathematical model for the building material panel type obtained in step 401 (for example, set by the production equipment operator), and whether there are sufficient training data sets for the mathematical model alone or in combination with other building material panel types.

[0228] In step 403, the control device 200 obtains process parameter selection information and / or information representative of at least one selected quality parameter, wherein the value of the quality parameter is predicted for the at least one building panel by the mathematical model. As described, the control device 200 may obtain the process parameter selection information and / or information on the selected quality parameter based in particular on user input.

[0229] Now, the control device 200 can generate a mathematical model based on the selection of process parameters and / or for at least one selected quality parameter. Figure 4 As shown, the method in the illustrated embodiment comprises a step 404 of obtaining information indicating the presence of at least one training data set that is not included in the number of training data sets on which the mathematical model is based.

[0230] This step may be performed immediately after step 402. For example, when switching production, step 402 may be used to determine whether a corresponding mathematical model has been used for the type of building material panel to be produced, for example, in a previous production cycle. The corresponding mathematical model (e.g., corresponding model coefficients, training data set, and / or data representing the corresponding mathematical equation) may be stored after the previous production cycle ends for subsequent use.

[0231] In step 404, it can now be determined that further sample data sets have been obtained in the meantime as training data sets, for example from a laboratory in which corresponding panel samples have been studied in the meantime, which training data sets are not yet available in the model finally used. However, step 404 can also (additionally or alternatively) be performed during the production process of manufacturing the at least one building material panel, for example when more sample data sets are available as training data sets during the production process.

[0232] In step 405, if at least one training data set contains and / or corresponds to a timestamp, wherein the timestamp represents a time point before a predetermined time interval (e.g., before a time interval before the start of the execution of the method and / or the manufacture of at least one building material panel), the control device discards the at least one training data set from the number of training data sets on which the mathematical model is based. Step 405 may be performed immediately after step 402 (before and / or after step 404). As described herein, the control device 200 may, for example, be configured to obtain information indicating that at least one training data set on which the mathematical model (e.g., a mathematical model used in a previous production cycle) is based contains or corresponds to a timestamp, wherein the timestamp represents a time point before a predetermined time interval. The control device 200 may, for example, obtain this information based on an input made through a user interface. As an alternative or in addition, the control device 200 may be configured to automatically determine, for example after step 402, whether at least one training data set on which the mathematical model is based contains or corresponds to a timestamp, wherein the timestamp represents a time point before a predetermined time interval. However, step 405 may also be performed at other points in time, for example during or after a production process for producing at least one building material panel.

[0233] like Figure 4As shown, method 400 also includes step 406, i.e., training and / or generating a mathematical model based on a training data set included in the number of training data sets (the number of training data sets in step 402), excluding the at least one discarded training data set (in step 405), and based on at least one training data set not included in the number of training data sets (according to step 404). In other words, steps 404 and 405 are update steps for identifying newly obtained training data sets and outdated training data sets. In step 406, a mathematical model is generated / trained based on the training data set that has not been discarded (already available) and the latest available training data set.

[0234] In step 407, the control device 200 prepares the mathematical model generated and / or trained in step 406 for use in the production of at least one building material panel. The control device can store the mathematical model generated in step 406 (corresponding model coefficients, corresponding training data sets, and / or data representing corresponding mathematical equations), for example, on the storage medium 250, for use in the production of at least one building material panel.

[0235] In step 408, the control device 200 acquires (e.g. calculates) prediction data based on the mathematical model and at least one sensory measurement value from a sensor of the production device 1 and / or at least one measurement value of the measuring device of the production device 1, wherein the prediction data represents a prediction of at least one quality parameter of at least one building material panel. The prediction data is then output in step 409. As mentioned above, the output prediction data can be converted into so-called control signals and / or so-called display data by the control device 200, which can be used by a display device (e.g. a screen connected to the control device) for displaying a representation or presentation.

[0236] Therefore, method 400 includes step 410, namely, based on the predicted data, prompting the display of a predicted representation of at least one quality parameter for the at least one building material panel, and / or generating at least one control signal based on the predicted data for controlling at least one unit and / or section of the production equipment 1 and / or for adjusting at least one process parameter of the production equipment 1.

[0237] Figure 5 An exemplary embodiment of a control device 200 is schematically shown, wherein the control device is configured to execute the method according to the aspect of the present invention. The control device 200 may be at least a part of a control device of a production facility, for example.

[0238] The control device 200 comprises: a processor 50 , a program memory 51 , a working memory 52 , a user data memory 200 , one or more communication interfaces 54 , a detection unit 55 , for example for detecting actual values ​​or set values ​​of one or more process parameters, and a user interface 56 .

[0239] The processor 50 executes, for example, a program stored in the program memory 51 to implement the method according to the aspect of the present invention. The working memory 52 is used in particular to store temporary data during the execution of the program.

[0240] The user data memory 250 is used to store data required during program execution and may correspond to Figure 1 The storage medium 250 in.

[0241] The communication interface 54 comprises one or more interfaces, in particular for communication between the device and the production device 1 and / or at least some parts of the production device 1 (e.g. one or more sensors and / or one or more measuring devices). The interface can be based on a wired connection and / or a wireless connection (e.g. a cellular mobile communication technology (e.g. GSM, E-GSM, UMTS, LTE, 5G) or a wireless local area network (WLAN)).

[0242] The user interface 56 can be designed as a screen and keyboard and / or a touch-sensitive display (touch screen). The user interface 56 can be directly (e.g. wired) connected to the processor 50 and / or the control device 200, and / or (multiple user interfaces 56 can be provided) connected to the processor 50 and / or the control device 200 via a wired and / or wireless (e.g. based on GSM, E-GSM, UMTS, LTE, 5G and / or WLAN (wireless local area network) technology) communication connection, such as a communication connection including an Internet connection. In the latter case, a remote connection to the control device 200 can be realized, which, for example, allows a user to remotely access the control device 200, thereby operating multiple control devices 200 when necessary.

[0243] The following exemplary embodiments are also part of the present disclosure:

[0244] Implementation Method 1

[0245] A method (100), for example performed by at least one device (200) or a system comprising at least two devices (200), the method (100) comprising:

[0246] - preparing (101) a mathematical model for use in the manufacture of at least one building material panel, wherein the mathematical model is based on a number of training data sets that is equal to or greater than a minimum number;

[0247] - wherein the training data set for a certain building material panel type comprises quality parameter data, the quality parameter data representing at least one value of at least one corresponding quality characteristic of a building material panel sample of the building material panel type, and process parameter data, the process parameter data representing values ​​of a plurality of process parameters of a production method for manufacturing the building material panel sample;

[0248] - wherein the number of training data sets includes at least one training data set of a building material panel type for the at least one building material panel, wherein the method further comprises:

[0249] - obtaining (102) prediction data based on the mathematical model, wherein the prediction data represents a prediction of at least one quality parameter of at least one building material panel;

[0250] - Output (103) prediction data.

[0251] Implementation Method 2

[0252] The method (100) according to embodiment 1, wherein the number of training data sets includes at least one training data set for a building material panel type of the at least one building material panel, and also includes at least one training data set for at least one other building material panel type.

[0253] Implementation 3

[0254] The method (100) according to embodiment 2, wherein a proportion of a production volume of building material panels of a building material panel type of the at least one building material panel in a total production volume is approximately equal to or less than a production proportion of building material panels of another building material panel type.

[0255] Implementation 4

[0256] A method (100) according to embodiment 2 or 3, wherein the mathematical model is based on the number of training data sets of the building material panel type of the at least one building material panel, and is also based on at least the number of training data sets of other building material panel types; wherein the number of training data sets of the building material panel type of the at least one building material panel is less than a minimum number and / or less than the number of building material panels of other building material panel types.

[0257] Implementation method 5

[0258] The method (100) according to any one of embodiments 2 to 4, wherein, for the building material panel type of the at least one building material panel and for at least one other building material panel type, a difference between at least one parameter value characterizing a property of the building material panel type of the at least one building material panel and at least one corresponding parameter value characterizing a corresponding property of the at least one other building material panel type is equal to or lower than a maximum value.

[0259] Implementation 6

[0260] The method (100) according to any one of embodiments 1 to 5, wherein the characteristics of the building material panel type are at least selected from:

[0261] - Thickness of building material panels;

[0262] - Width of building material panels;

[0263] - Building material panel density;

[0264] - Type of gluing of building panels;

[0265] - At least one material of the building material panel, in particular at least one glue type and / or wood material type.

[0266] Implementation 7

[0267] A method (100) according to any one of embodiments 1 to 6, wherein, for one or more quality parameters for which predictive data are obtained based on the mathematical model, the minimum number of training data sets is between 20 and 100, between 30 and 80, between 35 and 70, and / or between 40 and 60.

[0268] Implementation 8

[0269] The method (100) according to any one of embodiments 1 to 7, further comprising:

[0270] - obtaining information representing a building material panel type of the at least one building material panel;

[0271] - determining, based on the obtained information representing the type of building material panel of the at least one building material panel, whether there is a mathematical model used in the manufacture of the at least one building material panel, the mathematical model being based on a number of training data sets that is equal to or greater than a minimum number; and

[0272] - when it is determined that there is a mathematical model for use in the manufacture of the at least one building material panel, providing the mathematical model for use in the manufacture of the at least one building material panel.

[0273] Implementation method 9

[0274] The method (100) according to any one of embodiments 1 to 8, further comprising:

[0275] - training and / or generating a mathematical model for use in the manufacture of at least one building material panel based on a number of training data sets equal to or greater than a minimum number.

[0276] Embodiment 10

[0277] According to the method (100) of any one of embodiments 1 to 9, the training data set further has a timestamp and / or there is a corresponding association between the timestamp and the training data set, wherein the timestamp represents a time point at which the building material panel sample is produced, the method (100) further comprises:

[0278] if at least one training data set comprises and / or corresponds to a timestamp representing a time point located before a predetermined time interval, discarding the at least one training data set from the number of training data sets on which the mathematical model is based;

[0279] A mathematical model for use in the production of at least one panel of building material is trained and / or generated based on a number of training data sets that do not include the at least one discarded training data set.

[0280] Implementation 11

[0281] The method (100) according to any one of embodiments 1 to 10, further comprising:

[0282] - obtaining information representative of a statement that there is at least one training data set that is not included in the number of training data sets on which the mathematical model is based;

[0283] - training and / or generating the mathematical model based on the training data set included in the number of training data sets and based on the at least one training data set not included in the number of training data sets.

[0284] Implementation 12

[0285] The method (100) of any one of embodiments 1 to 11, wherein preparing the mathematical model for use in the production of the building material panel comprises:

[0286] Obtaining process parameter selection information representing a selection of process parameters for a production method for producing at least one building material panel and / or information representing at least one selected quality parameter, wherein a value of the at least one selected quality parameter is predicted for the at least one building material panel by a mathematical model;

[0287] The mathematical model is trained and / or generated based on the selection of the process parameters and / or for the at least one selected quality parameter.

[0288] Implementation 13

[0289] The method (100) according to any one of embodiments 1 to 12, wherein the method (100) is performed by at least one control device (100) of a production device (1) for producing building material panels or a system comprising such a control device (200), wherein the production device (1) has at least one sensor and / or measuring device, and the sensor and / or measuring device is configured to output at least one sensor measurement value and / or a measurement value of the measuring device as a target value or an actual value of a corresponding process parameter, and the process parameter characterizes a corresponding process condition of the production device during the production of the building material panel, wherein obtaining the prediction data based on the mathematical model further comprises:

[0290] - Prognostic data are acquired based on a mathematical model and based on at least one sensor measurement value and / or at least one measurement value of a measuring device.

[0291] Implementation 14

[0292] The method (100) according to any one of embodiments 1 to 13 further comprises:

[0293] - causing display of a predicted representation of at least one quality parameter for the at least one building material panel based on the predicted data; and / or

[0294] - generating at least one control signal based on the prediction data for controlling at least one component of the production device (1) and / or for adjusting at least one process parameter of the production device (1) when producing the at least one building material panel.

[0295] Implementation 15

[0296] A method (100) according to one of embodiments 1 to 14, wherein the mathematical model includes a simultaneous equation model, in particular a mathematical model based on a three-stage least squares method, a mathematical model based on a two-stage least squares method ("Two-stage leastsquares", "2SLS"), a mathematical model based on a partial least squares method (PLS), and / or a linear regression model.

[0297] Implementation 16

[0298] A device (200) or a system comprising at least two devices (200) is configured to execute and / or control a method (100) according to one of embodiments 1-15, or comprises corresponding devices for executing and / or controlling the steps of a method (100) according to one of embodiments 1-15.

[0299] Implementation 17

[0300] A computer program comprising program instructions which, when executed on a processor (50) or multiple processors (50), cause the one or more processors to execute and / or control the method (100) according to any one of embodiments 1-15.

[0301] The exemplary embodiments of the invention described in this specification are also to be understood as being disclosed in all combinations with one another. In particular, in this case, the description of a feature included in an embodiment is not to be understood as - unless expressly stated otherwise - that the feature is necessary or indispensable for the functioning of the embodiment. The order of the method steps described in this specification is not mandatory; unless otherwise stated, alternative orders of the process steps are contemplated. The method steps may be implemented in different ways and methods, for example, it may be contemplated to implement the method steps in software (via program instructions), hardware or a combination of both.

[0302] Terms such as "comprises", "having", "including", "containing", etc. used in the claims do not exclude further elements or steps. The wording "at least in part" includes both "partially" and "completely". The word "and / or" should be understood to disclose both alternatives and combinations, that is, "A and / or B" means "(A) or (B) or (A and B)". Multiple units etc. in this specification mean several units etc. The use of indefinite articles does not exclude plurality. A single device may perform the functions of multiple units or devices mentioned in the patent claims. The figure marks shown in the patent claims should not be regarded as limitations on the methods and steps used.

Claims

1. A method (100), for example performed by at least one device (200) or a system comprising at least two devices (200), the method (100) comprising: - preparing (101) a mathematical model for use in the manufacture of at least one building material panel, wherein the mathematical model is based on a number of training data sets equal to or greater than a minimum number; - wherein the training data set for a building material panel type comprises quality parameter data representing at least one value of at least one corresponding quality characteristic of a building material panel sample of the building material panel type, and process parameter data representing values ​​of a plurality of process parameters of a production method for manufacturing the building material panel sample; - wherein the number of training data sets includes at least one training data set for a building material panel type of the at least one building material panel, wherein the method further comprises: - obtaining (102) prediction data based on the mathematical model, wherein the prediction data represents a prediction of at least one quality parameter of at least one building material panel; - outputting (103) prediction data, wherein the number of training data sets includes at least one training data set for a building material panel type of the at least one building material panel, and at least one training data set for at least one other building material panel type.

2. The method (100) according to claim 1, wherein: The production ratio of building material panels of the building material panel type of the at least one building material panel in the total production volume is approximately equal to or less than the production ratio of building material panels of other building material panel types.

3. The method (100) according to claim 1 or 2, wherein the mathematical model is based on the number of training data sets of a building material panel type of the at least one building material panel, and is also based on at least the number of training data sets of other building material panel types; wherein the number of training data sets of a building material panel type of the at least one building material panel is less than a minimum number and / or less than the number of building material panels of the other building material panel types.

4. The method (100) according to any one of claims 1 to 3, wherein: For the building material panel type of the at least one building material panel and for the at least one other building material panel type, a difference between at least one parameter value characterizing a property of the building material panel type of the at least one building material panel and at least one corresponding parameter value characterizing a corresponding property of the at least one other building material panel type is equal to or lower than a maximum value.

5. The method (100) according to any one of claims 1 to 4, wherein the characteristics of the building material panel type are at least selected from: - Thickness of building material panels; - Width of building material panels; - Building material panel density; - Type of gluing of building panels; - At least one material of the building material panel, in particular at least one glue type and / or wood material type.

6. The method (100) according to any one of claims 1 to 5, wherein: For one or more quality parameters for which prediction data are obtained based on the mathematical model, the minimum number of training data sets is between 20 and 100, between 30 and 80, between 35 and 70, and / or between 40 and 60.

7. The method (100) according to any one of claims 1 to 6, further comprising: - obtaining information representative of a building material panel type of said at least one building material panel; - determining, based on the obtained information representing the type of building material panel of the at least one building material panel, whether there is a mathematical model used in the manufacture of the at least one building material panel, the mathematical model being based on a number of training data sets that is equal to or greater than a minimum number; and - when it is determined that there is a mathematical model for use in the manufacture of the at least one building material panel, providing the mathematical model for use in the manufacture of the at least one building material panel.

8. The method (100) according to any one of claims 1 to 7, further comprising: - training and / or generating a mathematical model for use in the manufacture of said at least one building material panel based on a number of training data sets equal to or greater than a minimum number.

9. The method (100) according to any one of claims 1 to 8, wherein: The training data set further has a timestamp and / or there is a corresponding association between the timestamp and the training data set, wherein the timestamp represents a time point at which the building material panel sample is produced. The method (100) further comprises: - if at least one training data set comprises and / or corresponds to a time stamp representing a time point located before a predetermined time interval, discarding said at least one training data set from the number of training data sets on which said mathematical model is based; - training and / or generating a mathematical model for use in the production of at least one panel of building material based on a number of training data sets excluding the at least one discarded training data set.

10. The method (100) according to any one of claims 1 to 9, further comprising: - obtaining information representative of a statement that there is at least one training data set that is not included in the number of training data sets on which the mathematical model is based; - training and / or generating the mathematical model based on the training data set included in the number of training data sets and based on the at least one training data set not included in the number of training data sets.

11. The method (100) according to any one of claims 1 to 10, wherein preparing a mathematical model for use in the production of building material panels comprises: - obtaining process parameter selection information representing a selection of process parameters of a production method for producing at least one building material panel and / or information representing at least one selected quality parameter, wherein a value of the at least one selected quality parameter is predicted for the at least one building material panel by means of a mathematical model; - training and / or generating a mathematical model in dependence on the selection of process parameters and / or for the at least one selected quality parameter.

12. The method (100) according to any one of claims 1 to 11, wherein: The method (100) is performed by at least one control device (100) of a production plant (1) for producing building material panels or a system comprising such a control device (200), wherein the production plant (1) has at least one sensor and / or measuring device, and the sensor and / or measuring device is configured to output at least one sensor measurement value and / or a measurement value of the measuring device as a target value or an actual value of a corresponding process parameter, wherein the process parameter characterizes a corresponding process condition of the production plant during the production of the building material panels, wherein obtaining prediction data based on the mathematical model further comprises: - Prognostic data are acquired based on a mathematical model and based on at least one sensor measurement value and / or at least one measurement value of a measuring device.

13. The method (100) according to any one of claims 1 to 12, further comprising: - causing display of a predicted representation of at least one quality parameter for the at least one building material panel based on the predicted data; and / or - In the production of the at least one building material panel, at least one control signal is generated based on the prediction data for controlling at least one component of the production device (1) and / or for adjusting at least one process parameter of the production device (1).

14. The method (100) according to any one of claims 1 to 13, wherein the mathematical model comprises a simultaneous equation model, in particular a mathematical model based on a three-stage least squares method, a mathematical model based on a two-stage least squares method ("Two-stageleast squares", "2SLS"), a mathematical model based on a partial least squares method (PLS), and / or a linear regression model.

15. An apparatus (200) or a system comprising at least two apparatuses (200), configured to execute and / or control the method (100) according to one of claims 1 to 14, or comprising corresponding apparatus for executing and / or controlling the steps of the method (100) according to one of claims 1 to 14.

16. Computer program comprising program instructions which, when executed on a processor (50) or a plurality of processors (50), cause the one or more processors to perform and / or control the method (100) according to any one of claims 1 to 15.

Images