METHOD FOR OUTPUTTING PREDICTION DATA OF A PREDICTION OF AT LEAST ONE QUALITY PARAMETER FOR AT LEAST ONE BUILDING MATERIAL BOARD

Description

Area

[0001] Exemplary embodiments of the invention relate to methods, devices, systems and computer programs, in particular for outputting prediction data of a prediction of at least one quality parameter for at least one building material panel. background

[0002] The publication by Mathias Jäckel et al., "Algorithm-based design of mechanical joining processes", PRODUCTION ENGINEERING, CARL HANSER VERLAG, DE, Vol. 16, No. 2-3, March 5, 2022 (2022-03-05), pages 285-293, XP037801088, ISSN: 0944-6524, DOI: 10.1007 / S11740-022-01121-2 discloses algorithm-based process models for certain self-pierce riveting processes ("mechanical joining process self-pierce riveting with semitubular rivet (SPR-ST)"), using process data for steel and aluminum joints for training and evaluating various prediction algorithms.

[0003] WO 2021 / 165394 A1 discloses the production of material panels, whereby a control system with an AI algorithm evaluates input parameters such as material, plant and product data and determines quality values. The control system verifies compliance with these values ​​and can simulate optimizations.

[0004] The continuous production of engineered wood and / or construction panels, such as wood-based panels, typically involves the use of complex production plants comprising multiple units or sections, each dedicated to specific production steps. For each such unit, a typically large number of process parameters can characterize the respective process conditions. For example, in the case of fiberboard production, process parameters such as the amount of squeezing water, the amount of wood chips, and the digester temperature can characterize the process conditions of a fiberization section within a wood-based panel production plant.

[0005] In this context, such process parameters influence the quality characteristics of building material panels produced under the corresponding production conditions, for example, transverse tensile strength, bulk density, flexural strength, etc.

[0006] Given the typically large number of process parameters that can characterize the process conditions of a production plant for manufacturing building material panels, it has proven advantageous to use mathematical models that describe a relationship between selected process parameters and resulting quality characteristics. Such models can help calculate the influence of set and / or measured target or actual values ​​of process parameters on expected quality characteristics, and, in the case of an operating production plant, predict corresponding quality characteristics based on available target or actual values ​​of process parameters.Corresponding predictive values ​​can be displayed to an operator based on available target or actual values ​​of process parameters, so that the operator of the production plant can react with appropriate settings of the production plant in order to adjust expected quality characteristics appropriately.

[0007] However, it has been shown that it is advantageous to create mathematical models for predicting quality characteristics specifically for a given type of panel to be produced. Due to the relatively large number of panel samples (also known as laboratory samples) required to generate a mathematical model for a given type of panel, this can lead to a situation where, in the case of panel types with a low share of total production, a corresponding mathematical model is only available after a comparatively long time, once enough sample data sets have been obtained.

[0008] It has also been found that, after a mathematical model has been created for a type of building material panel, predictions made by this mathematical model can worsen over time, for example depending on the season.

[0009] Against this background, a particular object of the present invention is to provide methods, devices, systems, and computer programs, especially for outputting predictive data representing a prediction of at least one quality parameter of at least one building material panel, even if the production share for that type of building material panel is small in relation to total production. Furthermore, it is an object of the present invention to provide methods, devices, systems, and computer programs, particularly for such types of building material panels, that output reliable predictive data. A further object of the present invention is to provide methods, devices, systems, and computer programs that support the control of a production plant for manufacturing building material panels. Summary of some exemplary embodiments of the invention

[0010] The object of the present invention is achieved by the subject matter of the independent claims. Advantageous embodiments are described in the dependent claims.

[0011] According to one aspect of the invention, a method for outputting predictive data of a prediction of at least one quality parameter for at least one building material panel is disclosed, wherein the method is carried out by at least one control device of a production plant for manufacturing building material panels, wherein the production plant has at least one sensor and / or a measuring device which is / are configured to output at least one sensor measurement value and / or a measurement value of the measuring device as a target or actual value of a corresponding process parameter which characterizes a corresponding process condition during the manufacture of the building material panel by the production plant, the method comprising: Providing a mathematical model for use in the manufacture of at least one building material panel, wherein the mathematical model comprises a set of model coefficients for one or more mathematical equations, wherein the one or more mathematical equations can output an expected value for at least one quality parameter based on an input of values ​​from at least one process parameter data set comprising values ​​for process parameters with associated timestamps, and wherein the mathematical model is based on a number of training data sets equal to or greater than a minimum number;wherein a training dataset for a building material board type comprises quality parameter data representing at least one value for at least one corresponding quality characteristic of a building material board sample of the building material board type, and process parameter data representing values ​​for a plurality of process parameters of a production process for manufacturing the building material board sample; wherein the number of training datasets comprises at least one training dataset for a building material board type of the at least one building material board, wherein the method further comprises: obtaining prediction data based on the mathematical model and based on at least one sensor measurement and / or at least one measurement from the measuring device, wherein the prediction data represent a prediction of at least one quality parameter for the at least one building material board; ; wherein obtaining the prediction data comprises: using the at least one obtained sensor measurement and / or the at least one obtained measurement from the measuring device as input or inputs for the mathematical model, and calculating the prediction data using the mathematical model based on the input or inputs; wherein the method further comprises: outputting the prediction data.

[0012] In accordance with this aspect of the invention, the following are further disclosed: A control device for a production plant for manufacturing building material panels comprising a processor, a program memory, a working memory, a user data memory, one or more communication interfaces, a data acquisition unit for acquiring actual or target values ​​for one or more process parameters, and a user interface, wherein the control device is configured to execute the method according to the aforementioned aspect of the invention; A production plant for manufacturing building material panels comprising at least one control device according to the aforementioned aspect, and a sensor and / or a measuring device, which is / are configured to output at least one sensor measurement value and / or a measurement value of the measuring device as a target or actual value of a corresponding process parameter.A computer program comprising program instructions that cause a control device according to the aforementioned aspect to execute the method according to the aforementioned aspect of the invention when the computer program is executed on the processor of the control device according to the aforementioned aspect. For the purposes of this specification, a processor shall include, but is not limited to, control units, microprocessors, microcontrollers, digital signal processors (DSPs), application-specific integrated circuits (ASICs), or field-programmable gate arrays (FPGAs). The process may either control all steps of the method, execute all steps of the method, or control one or more steps and execute one or more steps. The computer program may be at least partially software and / or firmware of a processor. It may also be implemented at least partially as hardware.The computer program can be stored, for example, on a computer-readable storage medium, such as a magnetic, electrical, optical, and / or other type of storage medium. The storage medium can be part of the processor, for example, a (non-volatile or volatile) program memory of the processor or a part thereof. The storage medium can also be a physical or tangible storage medium.

[0013] The following describes some characteristics of the aforementioned aspect – partly by way of example.

[0014] In exemplary embodiments, the method according to the aforementioned aspect of the invention can be carried out by at least one control device connected (directly and / or indirectly) to a production plant for manufacturing building material panels, in particular for controlling at least parts of a manufacturing process for building material panels. In an exemplary embodiment, the method can be carried out by a system comprising, for example, several control devices, each of which performs (some or all) steps of the method.For example, a control device may comprise a processing unit such as one (or more) computer that is (or are) provided at the production plant (and can, for example, receive sensor data and / or data from one or more measuring devices of the production plant via wired and / or wireless communication links) and / or a mobile device such as a tablet computer, a smartphone, or the like, which is configured to perform at least parts of the process and which is connected to the processing unit and / or the production plant, for example, via a wireless communication link. As further disclosed in this specification, the process according to the aforementioned aspect may be used at and / or to control at least parts of a production plant for manufacturing at least one building material panel, or the process may at least support such control.

[0015] The production of building material panels, sometimes simply called "material panels" in technical circles, takes place either in cycle-based or continuous processes. In cycle-based production, the building material panels are produced as flat objects with finite dimensions in all three spatial directions, whereas building material panels produced in a continuous process are lengths of a web material with finite dimensions in only two spatial directions. The operating principle of the joining and / or compaction unit determines whether the overall process is described as cycle-based or continuous, since in the case of cycle-based production, the process steps upstream of compaction, such as spreading, are often also designed as continuous subprocesses.Since significant pressures are generally applied in the compression units, or rather the combined joining and compression units, during the production of building material panels, these units are usually referred to by experts as the "pressing section" when referring to the entire system. In the production of building material panels as described in this document, the operating pressures, depending on the material and size of the panel being produced, are usually in the range of approximately 50 N / cm² to approximately 500 N / cm², and advantageously between 100 N / cm² and 400 N / cm², although they can be lower than 50 N / cm² in the production of insulating panels, so-called insulation panels, at very low densities.

[0016] Both economically and in terms of their technical applicability, building panels containing at least one layer of natural fibers occupy a special position among building panels. For the purposes of this document, natural fibers or fiber components are understood to be fibers and fiber components of natural origin, i.e., originating from an annual or perennial plant, regardless of whether the fibers or fiber components are present as pure fibers, for example, for the production of MDF / HDF panels or their layered versions, or as components of chips, long chips, or wafers, which are traditionally used for the production of particleboard or OSB panels or their layered versions. The term "wood particle," also used hereafter, therefore always includes at least natural fibers or fiber components.

[0017] Such building panels are often simply called "wood-based building panels" by experts, even if they have one or more layers that are not based on raw material derived from a perennial plant. Even building panels that comprise only one or more layers consisting at least partially of fibers and / or fiber components derived from annual plants are usually referred to as wood-based building panels and only very rarely, quite correctly, as "bast or grass-based building panels."

[0018] Wood-based building panels are manufactured in a wide variety of designs for different applications. Particularly widespread are particleboard, OSB panels, and MDF (medium-density fiberboard) or HDF (high-density fiberboard) panels, as well as hybrid panels constructed from individual layers of such composites. The naming of these building panels depends on the shape and size of the fibers or particles used in the panel or layer structure. A panel made from "fine" wood particles is called particleboard, while an OSB panel is made from "coarse" wood particles. Generally, "fine" wood particles are understood to be particles whose maximum dimension in any spatial direction does not exceed 60 mm; these particles, often described as chips, are usually formed with a maximum dimension of no more than 25 mm or even 20 mm.In general, experts understand "coarse" wood particles to be particles whose maximum dimension in any spatial direction is at least 60 mm; these particles, often described as long chips, are usually formed with a maximum dimension of 60 mm to 185 mm, and in particular from 80 mm to 140 mm.

[0019] MDF and HDF boards, or their individual layers, are made from (medium-density or high-density pressed) fibers, which are usually obtained from the raw material by means of an intermediate chemical process, usually a kind of cooking process.

[0020] Hybrid panels consist of several layers of different types and are often particularly suitable when the material panel has to meet various requirements for its intended use.

[0021] Such building panels are still referred to as wood-based building panels even if they consist of individual layers that contain no natural fibers or fiber components. These are usually laminated building panels, meaning wood-based building panels that are laminated on one or both sides. Plastics are most commonly used for lamination. So-called coated particleboard is a particularly well-known example.

[0022] The aforementioned types and varieties of wood-based building panels are therefore produced from wood particles (chips, long chips or fibers) of different shapes and sizes, whereby the wood particles are joined together in the so-called press section of a building panel production plant under the influence of pressure and temperature by stimulating their own adhesion mechanisms and adding adhesives (usually a glue).

[0023] More recently, efforts have been made to use annual plants, particularly grasses, for the production of building panels, in addition to wood-based materials that require many years to regenerate. These annuals have the significant advantage of rapid growth. Therefore, their use is particularly resource-efficient and aligns better with the growing global environmental awareness. Furthermore, increasing prosperity in many countries, especially in Asia, necessitates meeting a large demand for building panels for residential construction, particularly for interior finishing and furniture manufacturing.

[0024] Since annual plants do not develop bark, their harvested products initially form a homogeneous raw material from a production engineering perspective, the fibers of which can be obtained for the production of building material panels through a splicing process.

[0025] However, processing annual plants is significantly more complicated than processing wood particle-based building panels. The high precipitation of silicates during the manufacturing process, which are abrasive to the plant structure, presents a major obstacle. This necessitates considerably increased effort, particularly in plant construction, for example, through additional process steps, the reinforcement of certain plant components, and a higher demand for spare parts. Furthermore, there is a risk of production downtime.

[0026] Last but not least, the mechanical properties of plates (or layers) made from particles produced from annual plants differ considerably from those of counterparts based on wood particles.

[0027] The multitude of successive process steps, the variance of requirements for the individual process steps depending on the types of building material panels to be produced and their desired properties, as well as the parameters within a process step that affect the quality of the building material panel to be produced, result in almost arbitrarily complex dependency structures.

[0028] A building material from which the aforementioned building panels are manufactured comprises, in exemplary embodiments, a wood-based material, an insulating material, and / or one or more raw materials such as rice straw, bagasse, bamboo, hemp, or oil palm. Exemplary building panels include particleboard, oriented strand board (OSB), and / or MDF.

[0029] For the purposes of this disclosure, the term "building material panels" means panels that contain, for example, at least a significant proportion (e.g., more than 25 wt.%) of at least one of the aforementioned building materials. In the case of wood-based panels, for example, panels contain at least a significant proportion (e.g., more than 50 wt.%, in particular more than 75 wt.%, for example, at least 80 wt.%) of cellulose-containing material, such as lignocellulose-containing material, for example, wood chips. Such building material panels may include other materials, such as plastic materials, which may, for example, be present at least partially in particle and / or fiber form.

[0030] A production plant for manufacturing building material panels can therefore, in particular, comprise a production plant for manufacturing wood-based panels (or wood-substitute building material panels). Such production plants typically include several production units or sections, for example, a section (such as a bunker) for storage, debarking, and / or chipping; a section for washing wood chips; a section for fiberizing and / or machining and gluing; a section for fiber and / or chip drying; a section for mat forming; and / or a section for hot pressing.

[0031] Production conditions, particularly in individual sections and / or units of the production plant, can be characterized by corresponding process parameters. It is understood that process parameters, as defined in this disclosure, correspond in particular to actual values ​​and / or target values. Thus, in exemplary embodiments, the production plant includes at least one sensor and / or at least one measuring device. For example, the production plant may include one or more pressure sensors, temperature sensors, and / or humidity sensors. Furthermore, the production plant may include, for example, one or more measuring devices for measuring the properties of the building material panels, such as the properties of wood chips. However, it is understood that this disclosure is not limited to such sensors and / or measuring devices.In particular, the production plant has at least one production section that includes at least one sensor. For example, a press section can have a pressure sensor, where corresponding pressure readings can correspond to the process parameter pressure. Alternatively or additionally, a setpoint pressure at the corresponding press section can correspond to the process parameter pressure.

[0032] In other words, in an exemplary embodiment, the method is carried out by a control device of a production plant or by a system comprising such a 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 one measurement value of the at least one measuring device as a target or actual value of a corresponding process parameter, and which characterizes a corresponding process condition during the production of the building material panel by the production plant.

[0033] Exemplary process parameters, not limiting the present disclosure, particularly those relating to fiberizing and sizing, may include the amount of squeezing water, the amount of steam added to a cooker, the fill level of the cooker, the temperature of the cooker, the steam pressure of the cooker, the cooking time with the cooker, the amount of wood chips, the amount of paraffin added, etc. Exemplary process parameters, not limiting the present disclosure, particularly those relating to mat forming, may include the amount of fiber discharged, the spreading height, the forming belt speed, the basis weight of the mat, the mat moisture content, the spreading width, the pre-compression pressure, the pre-compression distance, the trimming width, the mat density, the amount of sprayed water, etc.It is understood that separate process parameters exist for other units and / or sections of a production plant, which characterize the corresponding production conditions of these other units and / or sections.

[0034] The production conditions prevailing / set in a production plant for manufacturing a building material panel have at least a partial influence on the properties of the produced building material panel. Such properties of a building material panel can be characterized, in particular, by quality features or quality parameters. For the purposes of this disclosure, quality features or quality parameters may include, in particular, transverse tensile strength, bulk density, flexural strength, and / or thickness swelling, and these examples are not to be understood as limiting. Thus, further quality features or quality parameters may be provided.

[0035] A production plant can be equipped with suitable sensors and / or measuring devices at various production stages, allowing production conditions during the manufacture of a building material panel to be recorded at each stage and assigned to a finished panel. Sensor data and / or measurement data from a measuring device, along with a corresponding timestamp, can be stored as a data set associated with a produced panel. These time-stamped data sets / data vectors contain the process conditions or states under which the respective building material panel was produced.

[0036] At regular intervals, typically produced building material panels, or parts thereof, are removed from the production process to measure their quality characteristics or parameters, for example, in a laboratory or using a test robot. This allows for the creation of sample data sets that include quality parameter data and corresponding process parameter data for a sample of a building material panel (whole panel or part thereof) of a specific panel type. These sample data sets thus provide a direct correlation between the set / prevailing process conditions at specific sections of the production plant, represented by the time-stamped process parameter data, and the resulting quality characteristics of the building material panel produced under these process conditions.Based on this information contained therein, the sample data sets can be used as training data sets for a mathematical model for use in the production of building material panels of the corresponding building material panel type, and are referred to below as training data sets.

[0037] Accordingly, according to the aforementioned aspect of the present invention, a training data set for a type of building material board comprises quality parameter data representing at least one value (e.g., a measured value obtained for a corresponding building material board sample in a laboratory) for at least one corresponding quality characteristic of a building material board sample of the building material board type, and process parameter data representing values ​​(or process parameter values, e.g., one or more target or actual values ​​for each process parameter, e.g., obtained by a corresponding sensor, measuring device, and / or setting device of the production plant) for a plurality of process parameters of a production process for manufacturing the building material board sample.

[0038] In exemplary embodiments, the process parameter data includes respective timestamps, or respective timestamps are assigned to the process parameter data (for example, stored in association with the process parameter data) that are associated with a corresponding process parameter. In other words, the process parameters of the majority of process parameters in exemplary embodiments can correspond to time-assigned process parameter values ​​that include a timestamp. The timestamp corresponds to the point in time at which a process parameter value characterizes a process condition that existed during the production of a manufactured building material panel.

[0039] In exemplary embodiments, the method comprises a step of providing (e.g. storing) at least one training data set on a storage medium that is connected to, or intended to be connected to, the production plant and that can be accessed by the at least one control device.

[0040] Based on the training datasets, it is possible to model the manufacturing process of a type of building material panel for a production plant. In exemplary embodiments, it may be possible, for instance, to assign corresponding coefficients or weights of a mathematical model to the process parameters, particularly for multiple training datasets, and to determine these in corresponding systems of equations within the mathematical model based on the measured quality characteristics. This can be done successively for ever more training datasets (for example, whenever new sample datasets are obtained from laboratory measurements during the production process), so that the mathematical model "learns" and can thus be "trained" with further training datasets.As mentioned, sample data sets used to train a mathematical model (on which the mathematical model is based) are accordingly referred to herein as training data sets.

[0041] As mentioned, the method according to the aforementioned aspect of the present invention comprises providing a mathematical model for use in the manufacture of at least one building material panel (for example, a series of several 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 for the building material panel type of the building material panel being manufactured and / or which is part of the series of building material panels being manufactured, which corresponds to a minimum number or which is greater than that minimum number.

[0042] In exemplary embodiments, a mathematical model comprises a set of model coefficients for one or more mathematical equations. In particular, in exemplary embodiments, a mathematical model comprises a set of model coefficients for one or more mathematical equations, wherein the one mathematical equation or the multiple mathematical equations with the model coefficients are configured such that, based on an input of values ​​from at least one process parameter dataset, which includes values ​​for process parameters with associated timestamps, the one mathematical equation or the multiple mathematical equations can output a value (an expected value) for at least one quality parameter.

[0043] In other words, for example, during the production of building material panels, set and / or measured target and / or actual values ​​of process parameters of the production plant can be transmitted to at least one control device, which is connected to the production plant via suitable communication links. Using the process parameter values ​​thus obtained, the at least one control device can calculate predictive values ​​for quality characteristics that can be expected for the building material panels currently being produced, based on the set and / or measured target and / or actual values, using the mathematical model.Accordingly, the procedure, as described in the above aspect, includes obtaining prediction data based on the mathematical model, wherein the prediction data represents a prediction of at least one quality parameter for at least one building material panel (the building material panel being manufactured and / or the panel being part of the series of building material panels being manufactured).

[0044] In one exemplary embodiment, obtaining the forecast data includes calculating the forecast data by and / or at the 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 another exemplary embodiment, obtaining the forecast data includes receiving the forecast data by the at least one control device, for example from a device connected to the control device at least via a communication link.

[0045] These predictive values ​​can be displayed to a production plant operator, allowing them to adjust process parameters and influence the expected quality characteristics if necessary. Additionally or alternatively, control signals can be generated based on the predictive data, enabling at least one control device to adjust process parameters for manufacturing at least one building material panel accordingly.

[0046] Thus, in an exemplary embodiment, making the mathematical model available 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 it on a storage medium that is connected to and / or accessible to the at least one production plant, selected from at least one of: at least one model coefficient for one or more mathematical equations; at least one training dataset for training and / or generating the mathematical model; data, for example program data, representing the mathematical equations.

[0047] In an exemplary embodiment, the training data sets (of the minimum number of training data sets) on which the mathematical model is based include training data sets that were used to train or generate the mathematical model for use in the manufacture of the at least one building material panel.

[0048] A mathematical model can be generated for use in the production of at least one building material panel, based on a number of training data sets available for the type of building material panel to be produced (stored on a storage medium connected to the at least one control device). This mathematical model is then used to predict at least one quality parameter of the building material panels to be produced and / or those already in production. In particular, if one or more building material panels are removed from ongoing production, measured in the laboratory, and thus new sample data sets become available as training data sets, the mathematical model can be updated based on these new training data sets.For this purpose, the mathematical model can be generated based on the previously existing training data sets and the new training data sets, which within the present disclosure is also understood as training the mathematical model based on the new training data sets.

[0049] Thus, in an exemplary embodiment, the method comprises: Generating and / or training the mathematical model for use in the manufacture of at least one building material panel based on the number of training data sets that is equal to or greater than the minimum number.

[0050] In exemplary embodiments, the mathematical model comprises or corresponds to a statistical model. In particular, in one exemplary embodiment, the mathematical model is based on a so-called model for interdependent simultaneous stochastic linear equations. It has been shown that, for example, regression models for interdependent systems of simultaneous (linear) stochastic equations can be suitable for modeling the production processes for building material panels. The processes in the wood-based panel industry, in particular, can be considered interdependent and simultaneous, since several product properties arise simultaneously and are not independent of one another, and there are mutual dependencies between the influencing factors or process parameters. The model equations can be described as linear, since the model coefficients are linear.It is understood, however, that the present disclosure is not limited to such linear systems. For example, process parameters can also be incorporated with a quadratic, cubic, or exponential (or other) weight.

[0051] Thus, in exemplary embodiments, the mathematical model comprises a simultaneous equation model, in particular a mathematical model based on the three-stage least squares method ("3SLS"). In exemplary embodiments, the mathematical model further comprises a mathematical model based on the two-stage least squares method ("2SLS"), a mathematical model based on partial least squares (PLS), and / or a linear regression model.

[0052] As mentioned, in one exemplary embodiment, the training data sets are stored (kept ready) on a storage unit for access by the at least one control device. In exemplary embodiments, the method according to the aforementioned aspect includes keeping the plurality of training data sets ready / stored, for example, on a storage medium to which the at least one control device is connected and / or which the at least one control device can access. In exemplary embodiments, a value of a process parameter represents a corresponding process condition of a production process for manufacturing a building material panel.

[0053] It is understood that data within the meaning of this disclosure, for example the quality parameter data, the process parameter data and / or data disclosed below, includes any type of information that can be read and / or processed by a processing device, such as the at least one control device, a processor, a system of one or more processors, and / or one or more computers, and which, in particular, can be stored for this purpose on a storage medium that is suitablely connected directly and / or indirectly (e.g., directly, wired or wirelessly) to the at least one control device, the processor, the system of one or more processors, and / or to the one or more computers.

[0054] As mentioned, the mathematical model is based on a number of training datasets that meets or exceeds a minimum requirement. This minimum requirement can be a number of training datasets large enough that training / generating the mathematical model with this number of datasets results in a model capable of generating predictive values ​​for at least one quality parameter, the accuracy of which meets a predetermined quality criterion. To verify whether predictive values ​​generated by a mathematical model meet the predetermined quality criterion, a predicted value of a quality criterion based on process parameter values ​​from a sample dataset can be compared with the actual measured value of the quality criterion from the sample dataset.For example, the quality criterion can be met if the absolute value of a difference between the predicted value and the actually measured value is less than or equal to a predetermined value.

[0055] In other words, in an exemplary embodiment, the minimum number corresponds to the number of training datasets for which a mathematical model, trained and / or generated based on this number of training datasets, can output a predictive value for at least one quality criterion, wherein a certain level of accuracy of the predictive value satisfies a predetermined quality criterion. In an exemplary embodiment, the minimum number of training datasets is between 20 and 100, between 30 and 80, between 35 and 70, and / or between 40 and 60 for the (one or more) quality parameters for which predictive data are obtained based on the mathematical model.

[0056] The minimum number can be fixed, for example, for a mathematical model of a building material panel type, or it can be flexibly adjustable, for example, by user input. The latter case, in particular, offers an operator the possibility to flexibly set the minimum number. In both cases, a value for the minimum number can be stored for access by the at least one control device on a storage medium that is connected to and / or accessible by the at least one control device. In other words, in an exemplary embodiment, the method comprises: Obtaining at least one piece of information representing a value of the minimum number, wherein the obtaining includes at least one of: receiving (e.g. receiving) the at least one piece of information based on an input via a user interface connected to the at least one device (the at least one control unit); accessing a storage unit connected to the at least one device (the at least one control unit) in order to obtain the at least one piece of information;

[0057] According to the aforementioned aspect of the present invention, the number of training data sets on which the mathematical model is based comprises at least one training data set for one (the) type of building material panel of the at least one building material panel (the building material panel that is manufactured and / or that is part of the series of building material panels that are manufactured).

[0058] For building material panel types that are frequently and / or in large quantities produced, and whose share of the total production volume is therefore correspondingly high, this minimum number can potentially be reached relatively quickly by training / generating the mathematical model based on training datasets for only one building material panel type—specifically, the type of panel that is frequently and / or in large quantities produced. Due to the high production volume and the correspondingly large number of panel samples that can be submitted to a laboratory for measurement, sample datasets can be generated at a relatively high frequency for such building material panel types, allowing the minimum number to be reached relatively quickly.

[0059] The present disclosure accordingly includes an exemplary embodiment in which the number of training data sets on which the mathematical model is based comprises a training data set only for one (the) type of building material panel of at least one building material panel (which is manufactured and / or which is part of the series of building material panels that are manufactured).

[0060] It is understood that, as long as the minimum number for a type of building material panel has not yet been reached, and a mathematical model for this type of building material panel cannot yet be used in the production of the corresponding building material panels for a prediction of expected quality parameters, the production of the type of building material panel can be carried out without such predictive values, based on the experience of the operators of the production plant.However, it has been found that controlling the production plant based on predictive values ​​by the mathematical model (for example, by an operator of the plant based on displayed predictive values ​​and / or by direct control of the plant based on control data generated based on the predictive data and used to directly / automatically / electronically control at least part of the production plant) leads to the production of building material panels with increased efficiency.

[0061] It is therefore desirable to be able to use a mathematical model even in the production of building material panels of a particular type, for which, for example, the minimum number of training data sets has not yet been reached due to a lower production share in the total output of the production plant. It has been found that it is possible to generate / train a mathematical model based on training data sets that are only partially for the building material panel type in question and partially for another type. In other words, in an exemplary embodiment, the number of training data sets includes at least one training data set for the building material panel type of the at least one panel, and at least one training data set for at least one other type of building material.

[0062] It has been found that this is particularly possible when the type of building material panel with the lower production share is grouped with a type of building material panel with a higher production share, i.e., when the number of training data sets consists of a number of training data sets for the type of building material panel with the lower production share and a number of training data sets for the type of building material panel with the higher production share. However, grouping is also possible, and often useful, when the production shares are approximately equal, differing, for example, by no more than 10%. Thus, in an exemplary embodiment, the production share of the total production quantity of building material panels of the type of building material panel of at least one building material panel is approximately equal to or less than the production share of building material panels of the other type.In other words, in an exemplary embodiment, the mathematical model is based on a number of training data sets for the type of building material panel of the at least one building material panel (which is manufactured and / or is part of the series of building material panels being manufactured), and on at least a number of training data sets for the other type of building material panel, wherein the number of training data sets for the type of building material panel of the at least one building material panel (which is manufactured and / or is part of the series of building material panels being manufactured) is less than the minimum number and / or less than the number of building material panels of the other type. Here, the total number of training data sets is equal to or greater than the minimum number, i.e.,The number of training records for the building material panel type of at least one building material panel (which is manufactured and / or is part of the series of building material panels that are manufactured) and the number of training records for the other building material panel type add up to a number that is equal to or greater than the minimum number.

[0063] It has been found that for a panel type for which there are not yet enough training data sets available (the number of which is less than the minimum required), grouping it with a panel type for which there are enough training data sets available (the number of which is equal to or greater than the minimum required) can be particularly advantageous if the panel types are similar in at least one property. Panel types can be characterized and / or classified, in particular, based on their properties, where, in exemplary embodiments, one property of a building material panel type is selected from at least: Thickness of building material panels; width of building material panels; density of building material panels; type of bonding of the building material panel; at least one material of the building material panel, in particular at least one type of adhesive and / or one type of wood material.

[0064] A building material panel can be characterized by one or more of these properties, wherein corresponding parameter values ​​for one or more of these properties are kept available for access by the at least one control device in exemplary embodiments, for example stored on a storage medium that is connected to the at least one control device and / or to which the at least one control device can access.

[0065] InIn an exemplary embodiment, the building material panel types are similar in at least one property if at least one parameter value representing at least one property is similar for the building material panel types. This can be the case if the difference between parameter values ​​for the two building material panels with respect to the at least one property does not exceed a maximum value, for example, a predefined value set by an operator.Thus, for the type of building material panel of at least one panel (which is manufactured and / or is part of the series of panels being manufactured) and for at least one other type of building material panel, a value representing the difference between at least one parameter value characterizing a property of the type of building material panel of at least one panel (which is manufactured and / or is part of the series of panels being manufactured) and at least one corresponding parameter value characterizing the corresponding property of the at least one other type of building material panel is equal to or less than a maximum value. In one exemplary embodiment, the value comprises an absolute value or a magnitude / absolute magnitude, although other values ​​that can characterize the difference between two parameters are also conceivable.

[0066] Since, as mentioned, it is advantageous to use a mathematical model based on a type of building material panel to be produced, it is also advantageous if a building material panel type, for example at the beginning of the production of a series of building material panels of a certain type, can be set for at least one control device. Accordingly, in an exemplary embodiment, the method comprises: Receiving (e.g. receiving) information that represents the type of building material panel of at least one building material panel (that is manufactured and / or that is part of the series of building material panels that are manufactured).

[0067] It is understood that information within the meaning of this disclosure, for example, the information representing the type of building material panel of the at least one building material panel and / or the information disclosed below, includes any kind of computer-readable and / or electronically readable information that can be read and / or processed by a processing device, such as the at least one control device, a processor, a system of one or more processors, and / or one or more computers.

[0068] In an exemplary embodiment, the information representing the type of building material panel of at least one building material panel (which is manufactured and / or is part of a series of building material panels being manufactured) is obtained based on input via a user interface. In exemplary embodiments, the user interface includes: at least one keyboard and at least one screen, one or more touchscreens, one or more means of voice input, and / or a mobile device connected to the production plant, for example a smartphone, a tablet computer, a laptop.

[0069] In exemplary embodiments, the user interface is directly connected – wired and / or wirelessly – to the production plant and / or to the at least one control device and / or, for example, via an internet connection. In exemplary embodiments, a 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 touchscreens. In exemplary embodiments, the at least one control device comprises a mobile device such as one or more smartphones, one or more tablet computers, and / or one or more laptops, wherein, particularly in these embodiments, a user interface comprises one or more touchscreens of a mobile device.

[0070] Based on the data representing the type of building material panel of at least one building material panel, the at least one control device can check whether a mathematical model exists for the type of building material panel to be produced (e.g., stored on one with the at least one control device) that is based on enough training sets (only for the type of building material panel to be produced or in grouping with another type of building material panel).

[0071] Accordingly, in an exemplary embodiment, the method comprises: Determine, based on the information obtained representing the type of building material of the at least one building material, whether a mathematical model exists for use in manufacturing the at least one building material, based on a number of training data sets equal to or greater than the minimum number.

[0072] It is understood that determining whether a mathematical model exists includes determining whether such a mathematical model (based on sufficient training datasets) is stored, for example, on a storage unit connected to and / or accessible by the at least one control device. It is further understood that connections, as defined in this disclosure, include in particular direct or indirect wired communication connections (e.g., LAN connections), and / or direct or indirect wireless communication connections, including radio connections such as Bluetooth, NFC, WLAN, 4G or 5G, and / or communication connections via the Internet. Indirect connections may include one or more connections via one or more intermediate nodes.

[0073] It is further understood that, in exemplary embodiments, storing a mathematical model includes storing at least one selected from: at least one model coefficient for one or more mathematical equations of the mathematical model; at least one training dataset for training and / or generating the mathematical model; data, for example program data, representing the mathematical equations.

[0074] If, based on the information obtained representing the type of building material of the at least one building material, it is determined that a mathematical model exists for use in manufacturing the at least one building material, based on a number of training data sets equal to or greater than the minimum number, the method in an exemplary embodiment further comprises: Having the mathematical model available for use in the manufacture of the at least one building material panel, if it is determined (has been determined) that the mathematical model is available for use in the manufacture of the at least one building material panel.

[0075] Otherwise, in an exemplary embodiment, no mathematical model is used in the production of the at least one building material panel.

[0076] In other words, in this exemplary embodiment, the step of making available a mathematical model for use in the manufacture of at least one building material panel includes making the mathematical model available for use in the manufacture of the at least one building material panel when it is (has been) determined that the mathematical model is available for use in the manufacture of the at least one building material panel.

[0077] As mentioned, a mathematical model developed for a type of building panel may, over time, produce predictions of reduced accuracy. Such changes can result from factors such as changing seasons, modifications to parts and / or materials of the production plant, or similar influences. However, it has proven advantageous to mitigate the impact of such long-term changes by no longer using older training datasets to generate / train a mathematical model for use in the production of at least one type of building panel.

[0078] In particular, for this purpose, an exemplary embodiment of a training data set includes a timestamp and / or an association exists between the timestamp and the training data set. The timestamp represents a point in time at which the building material panel sample was produced. For example, this point in time could be the point in time at which the building material panel sample was cut from a longer production line of building material panels and / or the point in time at which the building material panel sample was removed for measurement in the laboratory. It is understood that the timestamp represents such a point in time, but the exact numerical value is not important. For example, there could be a fixed time difference between the actual time of removal and / or cutting of the building material panel sample and the time value indicated by the timestamp.Such a timestamp makes it possible to discard training datasets when they are no longer up-to-date.

[0079] In an exemplary embodiment, the method further comprises a Obtaining information that represents a statement that at least one training dataset on which the mathematical model is based includes or is associated with a timestamp representing a point in time prior to a predetermined time interval.

[0080] This information can be obtained automatically from at least one control device and / or based on user input. Thus, in an exemplary embodiment, obtaining the information representing the statement that at least one training dataset on which the mathematical model is based includes or is associated with a timestamp representing a point in time prior to a predetermined time interval comprises at least one of the following: Obtaining the information representing the statement that at least one training dataset on which the mathematical model is based includes or is associated with a timestamp representing a point in time prior to a predetermined time interval, based on input via a user interface; Determining whether at least one training dataset on which the mathematical model is based includes or is associated with a timestamp representing a point in time prior to a predetermined time interval.

[0081] Accordingly, in an exemplary embodiment, the method further comprises: Discarding at least one training record from the number of training records on which the mathematical model is based if the at least one training record includes and / or is associated with a timestamp that represents a point in time prior to a predetermined time interval.

[0082] In an exemplary embodiment, the predetermined time interval comprises a time interval prior to the commencement of an execution of the method and / or the production of the at least one building material panel. In this exemplary embodiment, the time interval is defined in relation to the commencement of the execution of the method and / or the commencement of the production of the at least one building material panel. The time interval can, for example, be a number of years (e.g., 2 years) prior to the day on which the production of the at least one building material panel using the mathematical model is commenced. Thus, in an exemplary embodiment, the predetermined time interval comprises one or more days, one or more months, and / or one or more years prior to the commencement of the production of the at least one building material panel using the mathematical model.

[0083] In an exemplary embodiment, the method further comprises: Generating or training the mathematical model for use in manufacturing at least one building material panel based on a number of training datasets that do not include the discarded training dataset.

[0084] In an exemplary embodiment, making the mathematical model available for use in the manufacture of the at least one building material panel includes making the mathematical model available for use in the manufacture of the at least one building material panel, wherein the mathematical model is based on the number of training data sets that does not include the discarded training data set.

[0085] This approach advantageously allows older training datasets to no longer be used for training and / or generating the mathematical model, thus keeping the mathematical model up-to-date and mitigating long-term negative influences, such as those caused by seasons. In this way, the mathematical model can, in particular, produce more reliable forecast data.

[0086] As mentioned, a mathematical model for use in the production of at least one building material panel can be generated based on a number of training datasets available for the type of building material panel to be produced (stored on a storage medium connected to the at least one control device). In particular, if one or more building material panels are removed from ongoing production, measured in the laboratory, and thus new sample datasets become available as training datasets, the mathematical model can be updated based on these new training datasets. Likewise, the mathematical model can be updated, for example, before the start of production of the at least one building material panel, i.e., before the start of production of a series of building material panels of a particular type.

[0087] For such update processes, the method, in an exemplary embodiment, includes: Obtaining information that represents a statement that at least one training data set exists (e.g., is stored on a storage medium connected to and / or accessible by the at least one control device) that is not included in the number of training data sets on which the mathematical model is based.

[0088] In other words, the at least one control device in this embodiment is configured to obtain information that, in addition to the training data sets on which the mathematical model is based, further training data sets exist. In an exemplary embodiment, the step of obtaining the information that at least one training data set exists 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 the at least one building material panel using the mathematical model.

[0089] If information is obtained that at least one training data set exists that is not included in the number of training data sets on which the mathematical model is based, the method, in an exemplary embodiment, comprises: Training and / or generating the mathematical model based on the training datasets included in the number of training datasets, and based on at least one training dataset not included in the number of training datasets.

[0090] Thus, in this case, the mathematical model can be generated based on previously existing training datasets and new training datasets not yet included in the mathematical model. This can be done before the start of production of at least one building material panel and / or during production (during the manufacturing process) of the at least one building material panel. Within the scope of this disclosure, generating the mathematical model based on previously existing training datasets and newly added training datasets is also understood as training the mathematical model based on the new training datasets.

[0091] The information that at least one training data set exists that is not included in the number of training data sets on which the mathematical model is based can be obtained via input from a production plant operator through the aforementioned user interface and / or automatically detected by the at least one control device. Accordingly, in an exemplary embodiment, the method comprises: Obtaining information that represents a statement that at least one training dataset exists that is not included in the number of training datasets on which the mathematical model is based, based on input via a user interface.

[0092] Alternatively or additionally, obtaining the information representing the statement that at least one training dataset exists that is not included in the number of training datasets on which the mathematical model is based includes: Determine if at least one training dataset exists that is not included in the number of training datasets on which the mathematical model is based.

[0093] In this embodiment, training and / or generating the mathematical model based on the training datasets included in the number of training datasets, and based on the at least one training dataset not included in the number of training datasets, comprises: Training and / or generating the mathematical model based on the training datasets included in the number of training datasets, and based on the at least one training dataset not included in the number of training datasets, if it is determined that at least one training dataset exists that is not included in the number of training datasets on which the mathematical model is based.

[0094] In this exemplary embodiment, the method further comprises obtaining (for example, calculating by the at least one control device) prediction data based on the mathematical model, wherein the mathematical model is based on the training data sets included in the number of training data sets and on the at least one training data set not included in the number of training data sets, wherein the prediction data represent a prediction of at least one quality parameter for the at least one building material panel (the building material panel being manufactured and / or the panel being part of the series of building material panels being manufactured), and outputting the prediction data.

[0095] The described update process, based on the new training datasets, ensures that the model remains current and, in particular, is continuously improved to better adapt to a specific type of building panel within a production process. This is especially advantageous when a mathematical model is initially based on a grouping of training datasets for the building panel type to be produced with training datasets for a different type of building panel.In such cases, a continuous update of the mathematical model based on new training data sets obtained during the production process can make it possible that after a certain production time (possibly after several production cycles) enough training data sets are available for the type of building material board to be produced, so that predictions for this type of building material board become possible with a mathematical model that is based only on training data sets for this type of building material board to be produced.

[0096] A production plant for manufacturing building material panels typically comprises multiple production sections and / or units. The production conditions to which an intermediate product is exposed during the production of the building material panel at / within such a production section and / or at / within such a production unit can, in turn, be determined by a multitude of process parameters. Since, for example, corresponding sensors and / or measuring devices in a production plant determine corresponding target and actual values ​​for the respective process parameters during production, and these are stored in the sample data sets, at least for the respective building material panel samples (for which quality characteristics are measured in the laboratory), the sample data sets can contain target and actual values ​​for a large number (e.g., up to or even more than 3500 in a typical production plant) of process parameters.

[0097] Since not all process parameters necessarily have an equal influence on the resulting quality characteristics of a produced building material panel, it has proven advantageous to make a preselection of (typically several hundred, for example, 250 to 500) process parameters (to determine a group of process parameters) that is considered when adapting and / or generating the mathematical model (e.g., by the 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 allows, in particular, the computational effort used for adapting and / or generating the mathematical model to be controlled and, if necessary, minimized. It has also proven advantageous for the mathematical model to output predictive values ​​for a preselection of quality parameters.

[0098] Accordingly, having the mathematical model available for use in the manufacture of the building material panel in an exemplary embodiment includes: Obtaining process parameter selection information representing a selection of process parameters of the production process for manufacturing the at least one building material panel and / or information representing at least one selected quality parameter for which a value for the at least one building material panel (the building material panel being manufactured and / or the panel being part of the series of building material panels being manufactured) is to be predicted by the mathematical model; generating the mathematical model based on the selection of process parameters and / or for the at least one selected quality parameter.

[0099] In this process, the process parameter selection information and / or the information representing the at least one selected quality parameter can be obtained from 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 on the storage unit connected to the at least one control device. Alternatively or additionally, this information can be obtained via user input using the aforementioned user interface. In other words, in an exemplary embodiment, obtaining the process parameter selection information and / or the information representing the at least one selected quality parameter comprises at least one of the following: Obtaining the process parameter selection information and / or the information representing the at least one selected quality parameter based on input via a user interface; Obtaining the process parameter selection information and / or the information representing the at least one selected quality parameter based on a query of a database and / or a storage unit accessible to the at least one control device.

[0100] According to the aforementioned aspect of the invention, the method is carried out by at least one control device of a production plant for manufacturing building material panels or by a system comprising at least one such control device. The production plant comprises at least one sensor and / or a measuring device, which is configured to output at least one sensor reading and / or a measured value from the measuring device as a target or actual value of a corresponding process parameter that characterizes a corresponding process condition during the manufacturing of the building material panel by the production plant. Obtaining the predictive data based on the mathematical model further comprises: Obtaining the prediction data based on the mathematical model and based on at least one sensor measurement and / or at least one measurement from the measuring device.

[0101] According to the aforementioned aspect of the invention, obtaining the prediction data comprises using the at least one obtained sensor reading and / or the at least one obtained measurement value from the measuring device as input(s) for the mathematical model, and calculating the prediction data using the mathematical model based on the input value(s). Sensor readings and / or measurement values ​​from the measuring device can be received by the at least one control device, for example, via one or more corresponding communication links with at least one corresponding sensor and / or with at least one corresponding measuring device of the production plant. Thus, in an exemplary embodiment, the method comprises: Receiving at least one sensor measurement value and / or at least one measurement value from the measuring device via a communication link from at least one sensor of a production plant for manufacturing at least one building material panel and / or from at least one measuring device of the production plant for manufacturing at least one building material panel.

[0102] As mentioned, the method, according to the aspect mentioned, includes outputting the forecast data. As mentioned, the forecast data can be used to control a production plant for manufacturing the at least one building material panel and / or can at least support the control of the production plant. In an exemplary embodiment, the method thus further includes: controlling at least one unit and / or section of a production plant for manufacturing the at least one building material panel based on the output forecast data. It is understood that, in exemplary embodiments, controlling includes at least one of: direct control of the at least one unit and / or the at least one section based on corresponding control signals generated by the at least one control device based on the output forecast data, control of the at least one unit and / or the at least one section by an operator of the plant based on a display (e.g. a visual display by a display device) of the prediction of the at least one quality parameter based on the output forecast data.

[0103] Outputting the prediction data can therefore include internal output of this data, for example, within the at least one control device, for further processing of the prediction data by the at least one control device. Outputting the prediction data can also include outputting the prediction data for further processing by a device external to the at least one control device. For example, the output prediction data can be converted in the control device or in another data processing device into the aforementioned control signals and / or into display data that can be used by the aforementioned display device (e.g., a screen connected to the control device) to display a representation or representation.

[0104] In one exemplary embodiment, the method may include: Based on the output forecast data, cause a display of a representation of the prediction of at least one quality parameter for the at least one building material panel (the building material panel being manufactured and / or which is part of the series of building material panels being manufactured) based on the forecast data; and / or generate, based on the output forecast data, control signals to control at least one unit and / or section of the production plant and / or to set at least one process parameter of the production plant during the manufacture of the at least one building material panel.

[0105] In this context, a representation can, for example, include a graphical display of the forecast data (e.g., a real-time display) in the form of a data value-time curve on a screen, where the screen, as a display device, can be directly or indirectly connected to the control device and / or the production plant. Such a display of the forecast data can enable a user to adjust the production plant (e.g., in real time) in response to the displayed forecast values ​​of the quality characteristics.

[0106] In alternative or additional embodiments, the 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 on the basis of a feedback loop.

[0107] Further advantageous exemplary embodiments of the invention can be found in the following detailed description of some exemplary embodiments of the present invention, particularly in conjunction with the figures. However, the figures accompanying the application are intended only for illustrative purposes and not to limit the scope of protection of the invention. The accompanying drawings are not necessarily to scale and are intended only to reflect the general concept of the present invention by way of example. In particular, features included in the figures should by no means be considered a necessary component of the present invention.

[0108] They show: Fig. 1 is a schematic representation of an exemplary production plant for manufacturing a building material panel and a control device; Fig. 2 is an exemplary flowchart illustrating a method according to an exemplary embodiment of the aforementioned aspect of the invention; Fig. 3 is a schematic representation of a grouping of building material panel types; Fig. 4 is an exemplary flowchart illustrating steps of a method according to an exemplary embodiment of the aforementioned aspect of the invention; and Fig. 5 is a schematic representation of an exemplary embodiment of a device according to the aforementioned aspect of the invention, for example, a mobile device.

[0109] Fig. 1 shows a schematic representation of an exemplary production plant 1 for the manufacture of a material panel from chip material as an illustrative example of a building material panel in accordance with the present disclosure. Fig. 1Figure 1 further shows a schematic representation of a control device 200, which is connected to the production plant 1 for controlling the production plant 1, for example via the schematically depicted connection 550. For this purpose, the control device 200 can be configured to execute the steps of the method according to the aforementioned aspect of the invention. The control device 200 can, for example, comprise a processing unit such as a computer and / or a computer system, which is connected to (not shown) sensors and / or measuring devices of the production plant 1, for example to obtain target and / or actual values ​​of process parameters of sections or units of the production plant 1 from such sensors and / or measuring devices.Alternatively or additionally, in exemplary embodiments, the control device 200 can comprise one or more mobile devices, for example one or more smartphones, one or more tablet computers, and / or one or more laptops.

[0110] In exemplary embodiments, the control device 200 comprises one or more display devices, such as one or more screens, and / or is connected to one or more display devices to, for example, display at least one quality characteristic of a building material panel in production.

[0111] In exemplary embodiments, the control device 200 can comprise several control devices, one of which can perform one or more steps of the method according to the aforementioned aspect of the present invention. For example, it may be possible that the control device in Fig. 1 The control device 200, shown only schematically, comprises a processing unit, such as one or more computers, which is connected to sensors and / or measuring devices of the production plant 1 (wirelessly, wired, and / or via an internet connection), and, for example, a mobile device that is connected at least to the processing unit (wirelessly, wired, and / or via an internet connection). Thus, it is possible that, for example, in exemplary embodiments, steps involving the display and / or input of data and / or information may include the display and / or input of data using (for example, a touchscreen) a mobile device such as a smartphone.

[0112] Fig. 1further shows a storage device (for example, one or more hard drives and / or one or more cloud storage units) which may be intended, for example, for storing sample data sets and / or training data sets and is connected to the control device 200 via the schematically shown connection 500.

[0113] In exemplary embodiments, connections in accordance with the present disclosure include, in particular, the schematically illustrated connections 550 between the control device 200 and the production plant 1 (for example, between the control device 200 and one or more sensors, measuring devices and / or control controllers of the production plant 1) and 500, in particular, direct or indirect wired communication connections (for example, LAN connections), and / or direct or indirect wireless communication connections, including radio connections such as Bluetooth, NFC, WLAN, 4G or 5G and / or communication connections via the Internet.

[0114] Fig. 1Figure 1 shows a diagram with units, sections, or assemblies of production plant 1 that can be used in the manufacture of building material panels. In particular, the following sections of production plant 1 are shown, which can be used for individual production steps in the manufacture of a building material panel: a chipper 2, a chip dryer 3, a screening device 4, a device for applying a binder to the chips with mixers 5a, 5b, which also represent other devices with which binder is applied to the chips, furthermore, spreading devices 6, which spread the glued chips, if necessary in several layers of different chip sizes, from several spreading heads 6a, 6b onto a conveyor belt 7 to form a mat, a continuous press 8, and a cutting or milling device 9 for dimensioning the finished building material panels.

[0115] The chippers 2 also symbolically represent the general process of reducing wood to chips. The figure illustrates five knife-ring chippers, fed from a hopper above via screw conveyors. These can produce different chip sizes depending on the knife setting. The resulting chips, varying in size and moisture content, are then conveyed to the dryer 3.

[0116] The viewing device 4 shown in the figure can be used in different versions at various production points. This device can be used to separate fine particles (e.g., dust) or coarse particles (e.g., unwanted minerals or glue lumps) from the flow of chips. Viewing devices 4 can also be used, for example, to achieve fractionation according to chip size during the spreading process, so that different layers of the subsequent building material panel can be produced from chips of different sizes.

[0117] In the embodiment according to Fig. 1 For example, two different chip size ranges are produced, which are fed from hoppers to mixers 5a and 5b respectively. In these mixers, the chips are at least partially wetted with binder.

[0118] This makes it possible to supply different chip sizes to the spreading heads 6a for the top layers of the mat to be spread and pressed than to the spreading heads 6b for the middle layer(s).

[0119] The double belt press 8, as used for the production of building material panels, especially wood-based panels, has in its basic structure an upper press section with an upper heated pressure plate and a lower press section with a lower heated pressure plate. Frames, in which pressure transmitters for applying pressure are also supported, connect the upper and lower press sections. In both the upper and lower press sections, endless circulating steel belts are guided around belt deflection drums, forming a press gap for applying pressure and temperature to the mat.

[0120] Furthermore, one can recognize in Fig. 1A chip size measuring device 10 (an example of a measuring device) is also provided downstream of the mixer 5. Optionally, a chip size measuring device 11 can also be provided upstream of the mixer. A representative selection of chips, for example, a very small quantity of chips discharged from the transport process at a sampling point 20, can be measured in this chip size measuring device 10, 11. Exemplary process parameters that may be related to such chips measured by the chip size measuring device include, for example, the feed rate of binder via glue nozzles, the feed rate of chips, and / or the throughput speed of chips through the mixer 5a, 5b, the rotational speed of a shaft of a mixer 5a, 5b, etc.

[0121] A process according to the aforementioned aspect of the present disclosure may be used in connection with a production plant according to Fig. 1to be carried out, whereby it is understood that the present invention does not apply to the production plant according to Fig. 1 is restricted.

[0122] According to exemplary embodiments, process parameters can in particular be process parameters of wood-based panel production, for example process parameters of a fiberizing and gluing section of a production plant for the manufacture of wood-based panels, in particular comprising one or more process parameters that are selected from at least: Amount of squeezing water; steam addition to cooker; cooker fill level; cooker temperature; cooker steam pressure; cooking time; amount of wood chips; paraffin addition; energy consumption of refiner; temperature of refiner; steam pressure of refiner; grinding gap of refiner; grinding disc age; blow-off valve opening; pH value of the fibers; amount of glue.

[0123] In exemplary embodiments, process parameters of a mat forming section of a production plant for the manufacture of wood-based panels may include, in particular, one or more process parameters selected from at least: Fiber discharge rate; spreading height; forming belt speed; mat basis weight; mat moisture; spreading width; pre-compression pressures; pre-compression distances; trimming width; mat density; spray water quantity; mat height at forming belt end; mat temperature; spray height; misfill.

[0124] For such process parameters, appropriate sensors and / or measuring devices can be provided in a production plant, so that setpoint values ​​or actual values ​​corresponding to these process parameters can be made available to at least one control device via appropriate communication links.

[0125] In particular, it is possible, for example, to store measured and / or set target or actual values ​​during the production of a building material panel, along with corresponding timestamps, so that the relevant process parameter values ​​are available for a produced building material panel at the respective times. As described, for building material panel samples that can be sent out for laboratory analysis, for example, additional measurement values ​​for quality characteristics can be obtained. These can be stored together with the process parameter values ​​in a sample data set, or a training data set, for the corresponding building material panel sample.

[0126] According to exemplary embodiments, quality characteristics that can characterize a building material panel include, in particular, one or more quality characteristics selected from at least: Transverse tensile strength; bulk density; flexural strength; thickness swelling.

[0127] Thus, in exemplary embodiments, a sample data set or training data set for a building material board sample comprises data (process parameter data) that represent at least one value (e.g., target and / or actual value) for at least one corresponding process parameter with an associated timestamp, and data (quality parameter data) that represent at least one value (e.g., measured value and / or laboratory measured value) for at least one corresponding quality characteristic of a building material board sample.

[0128] Fig. 1 further shows a storage medium 250, which is connected to the control device 200 via connection 500, which the control device can thus access.

[0129] Fig. 2The flowchart shown is an exemplary flowchart representing an exemplary embodiment of method 100 according to the aforementioned aspect of the present invention. Flowchart 100 can be used to illustrate an exemplary control process for controlling a production plant for manufacturing a building material panel, for example, production plant 1 according to [reference to relevant invention]. Fig. 1 to be understood. Without limiting the invention thereto, it is assumed in the following that the method 100 is implemented by the control device 200 according to Fig. 1 is executed. However, in other exemplary implementation forms, method 100 can be executed by one or more processors of the control device 200, and / or by several control devices, wherein, for example, one or more processors and / or one or more of the control devices can execute one or more steps of method 100.

[0130] As in Fig. 2As described, Method 100 comprises a step 101 of storing 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. In other words, Method 100, as described herein, may include a step of storing, for example, the mathematical model (e.g., model coefficients for one or more mathematical equations, training data sets for training and / or generating the mathematical model, and / or data, such as program data, representing the mathematical equations).

[0131] As shown, a training data set for a building material board type includes quality parameter data that represent at least one value for at least one corresponding quality characteristic of a building material board sample of the building material board type, and process parameter data that represent values ​​for a plurality of process parameters of a production process for manufacturing the building material board sample.

[0132] As further explained, the number of training data sets includes at least one training data set for a type of building material panel of the at least one building material panel. As described herein, in exemplary embodiments, the number of training data sets may exclusively include training data sets for the type of building material panel of the at least one building material panel (the building material panel that is manufactured and / or that is part of the series of building material panels that are manufactured). InIn exemplary embodiments, the number of training data sets can comprise at least one training data set for the type of building material panel of the at least one building material panel (the panel being manufactured and / or part of the series of building material panels being manufactured), and at least one training data set for at least one other type of building material panel. In one exemplary embodiment, the mathematical model can be based on a number of training data sets for the type of building material panel of the at least one building material panel, and on at least a number of training data sets for the other type of building material panel, wherein the number of training data sets for the type of building material panel of the at least one building material panel is less than the minimum number and / or less than the number of building material panels of the other type.

[0133] This is exemplified in Fig. 3Illustrated for four different plate types A, B, C and D. Box 310 in Fig. 3 This illustrates a case where construction panel type A corresponds to at least one construction panel (the panel being manufactured and / or part of the series of panels being manufactured). As shown, sufficient training data is available for this panel type A to generate a mathematical model for use in the production of construction panels of type A based solely on training data for panel type A.

[0134] Box 330 illustrates an example case in which the number of training records for the building material panel type of at least one building material panel (the building material panel that is manufactured and / or that is part of the series of building material panels that are manufactured), in the case shown for building material panel type D, is less than the minimum number and less than the number of building material panels of the other building material panel type, in the case shown for building material panel type D.Since the properties of building material panels of types A and D are identical in the example case with regard to panel thickness and bonding type, grouping these two types allows for the creation of a mathematical model based on training datasets for panel type A and training datasets for panel type D. This model can then be used to predict the quality characteristics of panel type D, even though not enough sample datasets have yet been generated for this panel type alone. Using the described update processes, the mathematical model based on the grouping of training datasets for panel type D can be successively updated with training datasets for panel type D until grouping is no longer necessary for panel type D.

[0135] The grouping of building material panel type A and building material panel type D shown is advantageous for building material panel type D, for which very few training datasets are available, and is possible because sufficient training datasets are available for building material panel type A alone. Box 320 of the Fig. 3 In contrast, this shows a case of grouping building material panel types B and C, for which there are not enough training data sets available on their own. However, grouping these building material panel types B and C is still possible because the sum of the number of training data sets available for the individual building material panel types B and C is greater than the minimum number.

[0136] Again with reference to Fig. 2Method 100, as shown, comprises a step of obtaining prediction data based on the mathematical model, wherein the prediction data represent a prediction of at least one quality parameter for the at least one building material panel (the building material panel being manufactured and / or the panel being part of the series of building material panels being manufactured). As further described herein and as further in Fig. 2 As shown, the procedure further includes step 103 of outputting the forecast data.

[0137] Fig. 4 Figure 400 shows an exemplary flowchart illustrating the steps of a method according to an exemplary embodiment of the aforementioned aspect of the invention. Flowchart 400 can be used to illustrate an exemplary control process for controlling a production plant for manufacturing a building material panel, for example, production plant 1 according to Figure 1. Fig. 1to be understood. Without limiting the invention thereto, it is assumed in the following that the method 400 is implemented by the control device 200 according to Fig. 1 is executed. However, in other exemplary implementation forms, method 400 can be executed by one or more processors of the control device 200, and / or by several control devices, wherein, for example, one or more processors and / or one or more of the control devices can execute one or more steps of method 400.

[0138] As shown, procedure 400 includes a step 401 of obtaining information representing a type of building material panel of at least one building material panel to be produced. This step can be performed, for example, in connection with a production changeover when a type of building material panel to be produced is changed. The information regarding the building material panel type can be obtained, for example, based on input from a production plant operator via the aforementioned user interface.

[0139] As further described, the control device 200 determines, based on this, whether a mathematical model exists for the building material panel type that is based on a number of training data records equal to or greater than a minimum number. In other words, in step 402, the control device 200 checks whether there is a mathematical model for the building material panel type obtained in step 401 (for example, set by an operator of the production plant) for which, either alone or in combination with another building material panel type, there are sufficient training data records.

[0140] In step 403, the control device 200 receives process parameter selection information and / or information representing at least one selected quality parameter for which a value for at least one building material panel is to be predicted by the mathematical model. As described, the control device 200 can receive the process parameter selection information and / or the information regarding the selected quality parameter, in particular based on user input.

[0141] The control device 200 can now generate the mathematical model based on the selection of process parameters and / or for at least one selected quality parameter. As in Fig. 4As shown, the procedure in the example presented includes step 404 of obtaining information that represents a statement that at least one training data set exists that is not included in the number of training data sets on which the mathematical model is based.

[0142] This step can be performed directly after step 402. For example, during a production changeover using step 402, it may be determined that a mathematical model has already been used for the type of building material panel to be produced, perhaps in a previous production cycle. The corresponding mathematical model (e.g., relevant model coefficients, training datasets, and / or data representing relevant equations) may have been saved for further use after completion of the previous production cycle.

[0143] Step 404 can now determine that additional sample datasets have been obtained as training datasets in the meantime, for example, from a laboratory where corresponding plate samples have been examined, which were not yet available for the last used model. Step 404 can also be performed (additionally or alternatively) during the manufacturing process of producing at least one building material panel, if, for example, additional sample datasets become available as training datasets during the production process.

[0144] In step 405, the control device discards at least one training data set from the number of training data sets on which the mathematical model is based if the at least one training data set includes a timestamp and / or is associated with a timestamp representing a point in time prior to a predetermined time interval (for example, prior to the start of an execution of the process and / or the production of the at least one building material panel). This step 405 can be executed directly after step 402 (before and / or after step 404).As described herein, the control device 200 can, for example, be configured to receive information representing a statement that at least one training data set on which the mathematical model is based (for example, the mathematical model used in the previous production cycle) includes or is associated with a timestamp representing a point in time prior to a predetermined time interval. The control device 200 can obtain this information, for example, based on input via the aforementioned user interface. Alternatively or additionally, the control device 200 can be configured, for example, automatically after step 402, to determine whether at least one training data set on which the mathematical model is based includes or is associated with a timestamp representing a point in time prior to a predetermined time interval.However, step 405 can also be performed at another time, for example during or after the manufacturing process to produce at least one building material panel.

[0145] As in Fig. 4As depicted, procedure 400 further comprises step 406 of training and / or generating the mathematical model based on the training datasets included in the number of training datasets (the number of training datasets according to step 402), excluding the at least one training dataset discarded (in step 405), and based on the at least one training dataset (according to step 404) that is not included in the number of training datasets. In other words, steps 404 and 405 represent update steps in which newly available and outdated training datasets are identified. In step 406, the mathematical model is generated / trained based on the training datasets that were not discarded (which were already available) and the newly available training datasets.

[0146] In step 407, the control device 200 makes the mathematical model generated and / or trained in step 406 available for use in the production of the at least one building material panel. The control device can, for example, store the mathematical model generated in step 406 (corresponding model coefficients, corresponding training data sets, and / or data representing corresponding mathematical equations) on the storage medium 250 for use in the production of the at least one building material panel.

[0147] In step 408, the control device 200 receives (e.g., calculates) predictive data based on the mathematical model and based on at least one sensor measurement from a sensor of production plant 1 and / or at least one measurement from the measuring device of production plant 1, wherein the predictive data represents a prediction of at least one quality parameter for the at least one building material panel. The predictive data is output in step 409. As mentioned, the output predictive data can be converted by the control device 200 into the control signals and / or the display data, which can be used by the display device (e.g., a screen connected to the control device) to display a representation or representation.

[0148] Thus, the procedure 400 includes a step 410 of initiating a display of a representation of the prediction of at least one quality parameter for the at least one building material panel based on the prediction data and / or generating at least one control signal based on the prediction data to control at least one unit and / or section of the production plant 1 and / or to set at least one process parameter of the production plant 1.

[0149] Fig. 5 Figure 1 is a schematic representation of an exemplary embodiment of a control device 200 configured to execute the method according to the aforementioned aspect of the invention. The control device 200 can, for example, be at least part of a control device of a production plant.

[0150] 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 interface(s) 54, a data acquisition unit 55 for example for acquiring actual or target values ​​for one or more process parameters and a user interface 56.

[0151] The processor 50, for example, executes a program for carrying out the aforementioned method according to the aforementioned aspect of the invention, which is stored in the program memory 51. The main memory 52 serves in particular to store temporary data during the execution of this program.

[0152] The user data storage 250 is used to store data required for program execution and can be used with the storage medium 250. Figure 1 are equivalent to.

[0153] The communication interface(s) 54 comprise one or more interfaces for communication between the device, in particular with the production plant 1 and / or at least with parts (for example, with one or more sensors and / or with one or more measuring devices) of the production plant 1. The interface can be based on a wired and / or wireless connection (for example, on cellular mobile communications (e.g., GSM, E-GSM, UMTS, LTE, 5G) or on WLAN (Wireless Local Area Network)).

[0154] The user interface 56 can be configured as a screen and keyboard and / or as a touch-sensitive display (touchscreen). The user interface 56 can be directly connected (e.g., via a wired connection) to the processor 50 and / or the control device 200, and / or (multiple user interfaces 56 may be provided) it can be connected to the processor 50 and / or the control device 200 via a wired and / or wireless communication link (e.g., based on GSM, E-GSM, UMTS, LTE, 5G, and / or WLAN (Wireless Local Area Network) technology), for example, a communication link including an internet connection. In the latter case, a remote connection to the control device 200 can be enabled, allowing, for example, a user to access the control device 200 remotely and thus potentially operate multiple control devices 200.

[0155] The exemplary embodiments of the present invention described in this specification are to be understood as being disclosed in all combinations with one another. In particular, the description of a feature included in an embodiment—unless explicitly stated otherwise—is not to be understood as meaning that the feature is indispensable or essential for the function of the embodiment. The sequence of the process steps described in this specification is not mandatory; alternative sequences of process steps are conceivable—unless otherwise stated. The process steps can be implemented in various ways; for example, implementation in software (by program instructions), hardware, or a combination of both is conceivable.

[0156] Terms used in the claims, such as "comprise," "have," "include," "contain," and the like, do not exclude further elements or steps. The phrase "at least partially" covers both "partially" and "completely." The phrase "and / or" is to be understood as disclosing both the alternative and the combination; thus, "A and / or B" means "(A) or (B) or (A and B)." A plurality of units or the like, in the context of this specification, means several units or the like. The use of the indefinite article does not preclude a plurality. A single device can perform the functions of several units or devices mentioned in the claims. Reference numerals specified in the claims are not to be considered as limitations on the means and steps employed.

Claims

1. A method (100) for outputting prediction data of a prediction of at least one quality parameter for at least one building material board, the method (100) being carried out by at least one control device (100) of a production plant (1) for producing building material boards, - the production plant (1) having at least one sensor and / or one measuring device which is / are configured to output at least one sensor measurement value and / or one measured value of the measuring device as a setpoint or actual value of a corresponding process parameter which characterizes a corresponding process condition during the manufacture of the building material board by the production plant, the method (100) comprising: - holding available (101) a mathematical model for use in manufacturing at least one building material board, the mathematical model comprising a set of model coefficients for one or more mathematical equations, wherein the one or more mathematical equations can output an expected value for at least one quality parameter based on an input of values from at least one process parameter data set, which comprises values for process parameters with associated timestamp, and wherein the mathematical model is based on a number of training data sets which corresponds to a minimum number or which is greater than this minimum number; - wherein the minimum number corresponds to a number of training data sets for which the mathematical model can output a prediction value for at least one quality parameter when the mathematical model is trained and / or generated based on this number of training data sets, wherein a quality of the prediction value satisfies a predetermined quality criterion; - wherein a training data set for a building material board type comprises quality parameter data representing at least one value for at least one corresponding quality characteristic of a building material board sample of the building material board type, and process parameter data representing values for a plurality of process parameters of a production process for producing the building material board sample; - wherein the number of training data sets comprises at least one training data set for a building material board type of the at least one building material board, and at least one training data set for at least one other building material board type, the method further comprising: - obtaining (102) prediction data based on the mathematical model and based on at least one sensor measurement value and / or at least one measured value of the measuring device, the prediction data representing a prediction of at least one quality parameter for the at least one building material board; - wherein obtaining the prediction data comprises: - using the at least one obtained sensor measurement value and / or the at least one obtained measured value of the measuring device as an input variable or as input variables for the mathematical model, and - calculating the prediction data using the mathematical model based on the input variable or variables; - wherein the mathematical model is based on a number of training data sets for the building material board type of the at least one building material board, and on at least a number of training data sets for the other building material board type, wherein the number of training data sets for the building material board type of the at least one building material board is smaller than the minimum number; - wherein a sum of the number of training data sets for the building material board type of the at least one building material board and the number of training data sets for the other building material board type corresponds to a number that is equal to or greater than the minimum number; and the method further comprising: - outputting (103) the prediction data.

2. The method (100) according to claim 1, wherein a production share in a total production quantity of building material boards of the building material board type of the at least one building material board is equal to or less than a production share of building material boards of the other building material board type.

3. The method (100) according to one of claims 1 or 2, wherein the mathematical model is based on a number of training data sets for the building material board type of the at least one building material board, and at least a number of training data sets for the other building material board type, wherein the number of training data sets for the building material board type of the at least one building material board is smaller than the number of building material boards of the other building material board type.

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

5. The method (100) according to any one of claims 1 to 4, wherein a property of a building material board type is selected at least from: - building material board thickness; - building material board width; - building material board density; - type of gluing of the building material board; - at least one material of the building material board, in particular at least one type of glue and / or one type of wood material.

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

7. The method (100) according to any one of claims 1 to 6, further comprising: - obtaining information representing the building material board type of the at least one building material board; - determining, based on the obtained information representing the building material board type of the at least one building material board, whether there is a mathematical model for use in manufacturing the at least one building material board based on a number of training data sets equal to or greater than the minimum number.

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

9. The method (100) according to any one of claims 1 to 8, wherein a training data set further comprises a timestamp and / or there is an association between the timestamp and the training data set, wherein the timestamp represents a point in time at which the building material board sample was produced, the method (100) further comprising: - discarding at least one training dataset from the number of training datasets on which the mathematical model is based if the at least one training dataset comprises and / or is associated with a timestamp representing a point in time prior to a predetermined time interval; - training and / or generating the mathematical model for use in manufacturing the at least one building material board based on a number of training data sets which does not include 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 representing a statement that at least one training data set is present 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 sets 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 providing the mathematical model for use in manufacturing the building material board comprises: - obtaining process parameter selection information representing a selection of process parameters of the production method for manufacturing the at least one building material board and / or information re-representing at least one selected quality parameter for which a value for the at least one building material board is to be predicted by the mathematical model; - training and / or generating the mathematical model based 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, further comprising: - causing a display of a representation of the prediction of the at least one quality parameter for the at least one building material board based on the prediction data; and / or - generating at least one control signal based on the prediction data for controlling at least one component of the production plant (1) and / or for adjusting at least one process parameter of the production plant (1) during production of the at least one building material board.

13. The method (100) according to any one of claims 1 to 12, wherein the mathematical model comprises a simultaneous equation model, in particular a mathematical model based on the three-stage least squares method, a mathematical model based on the two-stage least squares method ("2SLS"), a mathematical model based on the partial least squares (PLS), and / or a linear regression model.

14. A control device (200) for a production plant (1) for manufacturing building material boards, comprising a processor (50), a program memory (51), a working memory (52), a user data memory, one or more communication interfaces (54), a detection unit (55) for detecting actual or setpoint values for one or more process parameters and a user interface (56), the control device (200) being set up for carrying out the method (100) according to one of claims 1-13.

15. A production plant (1) for producing building material boards, comprising at least one control device (200) according to claim 14, and a sensor and / or a measuring device which is / are configured to output at least one sensor measurement value and / or a measured value of the measuring device as a setpoint or actual value of a corresponding process parameter.

16. A computer program comprising program instructions that cause a control device (200) according to claim 14 to execute the method (100) according to any one of claims 1-13 when the computer program is executed on the processor (50) of the control device (200) according to claim 14.