Process for preparing a raw material, process for manufacturing a workpiece, raw material processing plant, processing plant

Abstract

The invention relates to a method for processing a raw material (2) by means of a processing plant (1), comprising the process steps: A) Providing raw material (2) to the processing plant (1), wherein the raw material (2) has a variety of different components; B) Division of the raw material (2) into a plurality of fractions (10, 11, 12); C) Feeding the fractions (10, 11, 12) into a plurality of cleaning devices (18, 19, 20); D) Cleaning the fractions (10, 11, 12) in the cleaning devices (18, 19, 20), whereby a quantity of contaminant is removed from the fractions (10, 11, 12) as contaminant fraction (22, 23, 24); E) Spreading the fractions (10', 11', 12') as prepared starting material (48). It is essential that at least one material-related process parameter is recorded and evaluated using an artificial intelligence-based algorithm in at least one of the process steps A), B), C), D) or E), and that at least one control variable of one of the process steps A), B), C), D) or E) is changed depending on a processing target variable, which depends at least on a throughput of the processed starting material (48) and / or a cleaning efficiency of the processing plant (1). The invention further relates to a method for producing a plate-shaped workpiece, a processing plant (1), and a processing plant.

Description

The invention relates to a method for processing a raw material according to the preamble of claim 1, a method for producing a workpiece according to claim 12, a processing plant for raw material according to the preamble of claim 13 and a processing plant according to claim 14.Raw material processing may be required for the production of numerous products in order to be able to achieve or improve the desired properties of said products. Here, in particular, impurities present in the raw material are discharged from the raw material. The interfering substances typically include foreign materials which have to be separated from the raw material, for example, on account of an undesired substance composition, geometry or size. Such processing can be effected in particular in the production of components from composite materials which can consist of biogenic materials, i.e. materials derived from natural biological sources, such as wood, or else of synthetically produced materials, such as plastic. In particular, the processing can take place in the context of recycling, i.e. recycling, of previously circulated products.DE 10 2021 118 108 A1 discloses a method for producing recycled scrap, wherein constituents are separated out from a comminuted material mixture on the basis of detected material information.Typically, the processing of raw material comprises a plurality of method steps, in which the raw material is first provided in a processing plant, wherein the raw material has a plurality of parts of different sizes. The raw material is divided into a plurality of fractions. The fractions are supplied to a plurality of purifying devices and the fractions are purified, wherein a quantity of impurity is removed in each case. Subsequently, the raw material in the fractions is discharged as a processed raw material and can be further processed. The fractions can subsequently be merged again, but often remain separated and can be treated further separately from one another.One challenge in performing the steps described above is that the raw material may have a very heterogeneous composition. In particular, wood or other naturally occurring materials can differ greatly from supplier to supplier depending on the origin and preparation. This has the result that the properties of the raw material which is located in the processing plant can change greatly within the operating period of the processing plant, often within a few minutes or even several hours. In other words, the fractions of a first raw material can thus be supplied to the cleaning devices in a ratio at which the cleaning devices are indeed used in the desired manner. If, however, a second raw material is subsequently to be processed, which is different from the first raw material, but whose fractions are formed according to the same criteria, this can lead to individual cleaning devices being overloaded and others not being used up sufficiently. This leads to an overall reduced plant efficiency.It is therefore an object of the invention to increase the efficiency in the processing of raw material.The object is achieved by means of a method according to claim 1 and by means of a processing plant according to claim 13. Furthermore, the object is achieved by a production method according to claim 12 and a processing plant according to claim 14.The method for processing raw material comprises in a manner known per se the following process steps: A) providing raw material in the processing plant, wherein the raw material has a plurality of different constituents; B) dividing the raw material into a plurality of fractions; C) feeding the fractions into a plurality of cleaning devices; D) cleaning the fractions in the cleaning devices, wherein in each case a quantity of interfering substance is removed from the fractions as an amount fraction of interfering substance; E) discharging the processed fractions as a processed starting material.It is essential that at least one material-related process parameter is detected at least in one of the process steps A), B), C), D) or E) and evaluated by means of an algorithm based on an artificial intelligence, and that at least one manipulated variable of one of the process steps A), B), C), D) or E) is changed depending on a treatment target variable which is dependent at least on a throughput of the treated starting material and / or a cleaning efficiency of the treatment plant.The invention is based on the finding that the use of artificial intelligence makes it possible to take account of material-related process parameters and to control at least one of the process steps during the execution of the processing method in order to be able to improve the efficiency of the plant. The invention is likewise based on the finding that the throughput of the processed starting material and the cleaning efficiency can serve as target variables, on which the plant efficiency is critically dependent, and that these variables are suitable for being input into the algorithm as target variables and for being able to control the above-mentioned process steps. It is thus an advantage of the method according to the invention that the method is not only controlled in order to maximize the throughput and the cleaning efficiency. Rather, it is possible to specify such target variables as required, so that the process steps A), B), C), D) or E) can be controlled taking into account at least one material-related process parameter.In particular, the starting material can be used for producing workpieces, preferably panel-shaped workpieces, particularly preferably wood material panels.According to the invention, it can be provided to train a model for the algorithm in such a way that takes into account a complex relationship between at least one material-related process parameter and at least one manipulated variable of the process steps A), B), C), D) or E) and the throughput of the processed starting material and / or a cleaning efficiency of the processing plant. In contrast to an analytical model, by using an algorithm based on an artificial intelligence, circumstances in the operation of the preparation plant that are not influenced by human assumptions during the modeling that may be erroneous can thus be taken into account. This is particularly relevant in the present case since the throughput and / or the cleaning efficiency of the plant are economic characteristic variables which are not directly related to the technical variables of the treatment plant or the raw material, so that, for example, an analytical model would be particularly susceptible to errors. Nevertheless, a surprising finding is that the algorithm can be designed in such a way that the technical components of a processing plant can be controlled with the aid of the algorithm as a function of predefinable economic characteristic variables.An artificial intelligence-based algorithm is understood in particular to mean an algorithm which is based on methods which are used in the development of artificial intelligence. Advantageously, an algorithm based on artificial intelligence comprises or is based on at least one method or a modification thereof from the following groups: simple methods, such as linear regression, polynomial regression, functional regression; and / or advanced methods such as, for example, random forest regression, support vector regression, K nearest neighbor regression, neural networks, recurrent neural networks, convolutional neural networks, residual networks, Bayesian networks and / or classification-based methods such as, for example, K nearest neighbor classification, decision trees, random forests, naive bayes, support vector machines.A simple method, such as linear regression, polynomial regression, functional regression, is understood to mean, in particular, widely used methods which originate from the field of statistics and can be used for the purposes of machine learning.An advanced method, such as, for example, random forest regression, support vector regression, K nearest neighbor regression, neural networks, recurrent neural networks, convolutional neural networks, residual networks, Bayesian networks, is understood in particular to mean methods which fall almost exclusively within the range of those of machine learning and have a lower meaning outside this domain. Classification-based methods such as, for example, K nearest neighbor classification, decision trees, random forests, naive bayes, support vector machines can also be regarded as a subgroup of the advanced methods.In addition to a mathematical method, in particular with mapping functions and / or adjacent functions, limits for specific parameters are preferably predefined for the algorithm, in which the algorithm can move freely. For a method based on neural networks, one or more mapping layers and weighting factors are predefined, for example.Preferably, an artificial neural network is used as an artificial intelligence-based algorithm. The use of an artificial neural network is understood to mean the use of machine learning in which no model is continuously predefined on the basis of physical values whose model parameters have to be adapted, but rather the use of an information processing system which is typically constructed from layers and can be defined by training methods. In an upstream design step, therefore, no model formation takes place in the actual sense. Rather, only a frame structure is specified, the parameters of which are optimized by means of an optimization algorithm. One or more layers of an artificial neural network can thereby be defined and / or created and, in particular, starting values for the weighting parameters from the past experiences can be taken into account. The above-mentioned optimization algorithm requires or does not have any information about the process which is simulated and merely adjusts all network parameters such that, in the case of the training data, the output variables coincide as well as possible with the predefined target variables.Within the scope of the invention, the raw material can be biogenic material, in particular wood material, as mentioned at the beginning. The method according to the invention can be carried out in particular for processing raw material for the production of workpieces or however for producing products which are processed further or used in some other way, for example as process engineering auxiliaries.In particular, the prepared starting material can be provided and used for the production of plate-shaped workpieces, which preferably include fiber boards, particle boards or oriented strand board (OSB) boards. These plate-shaped workpieces are therefore particularly relevant since their raw material can typically have a multiplicity of parts of different sizes or structures, for example in the form of fibers, chips, wood chips or pieces of wood. Furthermore, the raw material can contain straw, cereal husks, corn husks, plastic, paper or a foam, in particular polyurethane foam.In addition to plate-shaped workpieces, the prepared starting material can also be used for non-plate-shaped workpieces. These differ from plate-shaped workpieces in that they have larger dimensions transversely to a main extension plane and can be designed in particular as solid parts. These include, for example, pallet blocks. The prepared raw material can also serve for producing pallets or also three-dimensionally compression-molded workpieces.Furthermore, the prepared starting material can be further processed by means of a subsequent process. This is conceivable in particular if the raw material is not used for producing a workpiece, but is to be used as a starting product for producing a further product, for example in the production of bioethanol or methane. The processed starting material can also be used in another way, for example for converting thermal energy into electrical energy in combustion installations.Within the scope of the invention, a fraction can be considered to be a part of the raw material which differs from another part of the raw material, i.e. another fraction, by a characteristic feature. When the raw material is divided, it can therefore be grouped, for example, as a function of previously set size classes: if the raw material has, for example, a plurality of constituent parts of different sizes, these are typically present with an individual frequency distribution. In the fractionation, limits are defined for this frequency distribution, to which limits the total amount of raw material is divided accordingly. This results in several fractions in this example, in which the raw material is grouped according to size. It is within the scope of the invention that the fractionation can take place as a function of a fundamentally arbitrary property of the raw material, which can in particular also be a material-related process parameter which is detected.It is within the scope of the invention that the raw material is present without further mechanical or chemical treatment with a distribution that it can be divided into the desired fractions. However, it is also possible for the raw material or parts thereof to be preprocessed before method step B) in order to be able to adapt the distribution as required. This can be done, for example, by comminuting a quantity of coarse-grained raw material to a quantity of fine-grained material and feeding it back to the treatment plant.The fractionation can in principle be carried out using any fractionation apparatus. The fractionating device may be, for example, a screening device in which the raw material is fed along a feeding direction onto a screen surface having a plurality of screen openings of different sizes, and pieces of the raw material are retained on the screen surface or separated from the remaining raw material depending on the position. Below the screen surface, collecting devices can be arranged, by means of which the fractions can be collected spatially separated from one another. By positioning the collecting devices or their feed openings, it is thus possible to set the extent to which the fractions are to have in each case, that is to say the size to which a part of the raw material which is assigned to a fraction is to have.In a simple example, the raw material is divided into at least two fractions, a first fraction containing a first portion of the raw material whose parts have a size that is below a set limit value and a second fraction containing a second portion of the raw material whose parts have a size that is above the set limit value. The fractions may have an identical scope of parts or may be different. The limit value can preferably be defined by means of the algorithm on the basis of the artificial intelligence and in particular be dynamically adapted.The fractions are fed to the cleaning devices, by means of which they are cleaned and the impurities are removed by freezing. In particular, the cleaning devices per fraction may comprise one or more cleaning machines / devices, which may be considered cooperatively as a cleaning device. The interfering substances include, for example, minerals, glass, ferrous metals, non-ferrous metals, organic materials, such as straw, plastics or another substantially freely definable material. The cleaning can be carried out in fundamentally any desired manner, but preferably magnetically, in particular by means of a magnetic separator, inductively, in particular by means of an eddy current separator, mechanically, in particular by means of a separating screen, pneumatically, in particular by means of compressed air blows or wind classification, or chemically, for example by means of a solvent, with which the raw material is permeated. It is also within the scope of the invention that a sensor-based detection of the interfering substances takes place so that they can be removed from the load. In particular, an acoustic method, an optical method, preferably by means of color and / or shape recognition, a thermographic method, a radiographic method, in particular by means of X-ray radiation or an electromagnetic method, in particular by means of near-infrared radiation, can be used for this purpose. It is also within the scope of the invention that the mentioned cleanings, in particular the sensor-based cleanings, can themselves be KL-controlled. Correspondingly Kl-controlled cleaning devices / machines can be incorporated into the superordinate controller of the processing plant or can be operated separately therefrom.The fractions are then discharged as processed starting material, wherein they can be spatially separated or combined.The cleaning efficiency can be considered as a characteristic variable during operation of the processing plant, which describes a ratio between the achieved purity level of the discharged starting material and the resource outlay required for this purpose. The throughput can likewise be regarded as a characteristic variable during operation of the processing plant, which describes a quantity of the output starting material or input material introduced, i.e. of the raw material, per unit of time.Investigations by the applicant have shown that the material-related process parameters during process steps A), B), C), D) or E) have a decisive influence on the treatment target variables and must be taken into account for the plant operation in order to be able to control or regulate process steps A), B), C), D) or E) by changing the manipulated variables. The invention is in principle not restricted to which material-related process parameters are detected in process steps A), B), C), D) or E). Rather, a distinct realization is that the algorithm based on the artificial intelligence is capable of taking into account a material-related process parameter from any process step A), B), C), D) or E) in order to generate at least one proposal as to which manipulated variable would have to be changed in one of the process steps A), B), C), D) or E) in order to achieve the desired treatment target variable. It is therefore particularly within the scope of the invention that a material-related process parameter is determined in one of the process steps by measurement or testing and a control variable of the same and / or of another process step is changed.The material-related process parameter can be a property of the raw material, one of the fractions or the starting material which are determined during the execution of one of the process steps A), B), C), D) or E). In particular, this is the size, in particular a length or thickness, of a raw material constituent, temperature, moisture, geometry, speed, density, weight, material type or a comparable property of the raw material, which can be detected and, depending on which property, the raw material can be fractionated in principle.Preferably, in process step A), at least one material property of the raw material, preferably a ratio between a fines fraction and a coarse fraction and / or a moisture and / or a length dimension, is detected as a first material-related process parameter. In the case where the raw material comprises wood, a ratio between a solid wood fraction and a residual wood fraction can also be detected as the first material-related process parameter.The above-mentioned material properties of the raw material can be determined, for example, by means of an optical measuring device. For this purpose, the processing system can have one or more cameras which are connected by signal technology to a control unit and which capture the optical feature and output it to the algorithm. The control unit can be integrated into the optical measuring device and / or be a component of a higher-order system controller. Accordingly, the algorithm may be partially implemented in the optical measurement device and / or in the plant controller. The humidity can be determined by means of a humidity sensor directly on the raw material or directly via an air humidity that is established in the environment of the raw material.Additionally or alternatively, in process step B) at least one optical feature of at least one of the fractions or of a fraction of the raw material can be detected as a second material-related process parameter.An optical characteristic of the fractions can be, for example, an amount of material and / or a density of material, which can likewise be determined by means of an optical measuring device. If a screen is used for dividing the raw material, the optical measuring device can be arranged above the screen surface. This measuring device can be configured to determine which amounts of the raw material are retained or separated from the raw material. With this information it is possible to conclude whether a certain size class is over- or under-represented in the fractionation due to the nature of the raw material, which could lead to an uneven utilization of the cleaning devices. It is conceivable in principle that the measuring device is integrated in any desired component of the processing plant.Furthermore, additionally or alternatively, in process step C) at least one throughput of the fractions can be detected as a third process parameter.The flow rate of the fractions corresponds to a flow rate of a fractionated raw material per unit time. In other words, this is a stream of raw material of a size class which is fed to a purification device after fractionation. The throughput can be determined by means of a suitable measuring device, wherein in particular a gravimetric and / or a volumetric and / or radiation-related measuring device, in particular an X-ray device, can be used.In process step D), at least one unfrired interference quantity fraction or a purity level of at least one of the fractions can preferably be detected as a fourth process parameter.The fraction of interfering quantity corresponds to a fraction of interfering substance which is separated from a remaining crop fraction. The portion of interfering quantity may preferably be present here as a relative parameter which gives the quantity of separated interfering substance as a percentage relative to the remaining quantity of interfering substance. The portion of interfering quantity is preferably determined on the basis of the amount of interfering substance removed from a cleaning device. In addition, the supplied quantity of raw material of a fraction or optionally the discharged quantity of processed raw material can be taken into account, wherein the respective quantity of interfering substance is determined before or after the supply of the fraction into the cleaning device, in order to be able to determine the interfering quantity proportion, for example, by forming a difference. The purity level can correspond to a characteristic quantity which specifies a ratio between the processed starting material of a fraction and an amount of interfering material which has not dissolved out.In process step E), preferably at least one throughput of the processed starting material of a fraction can be detected as a fifth process parameter.The throughput of the processed starting material of a fraction corresponds to a delivery rate of the processed starting material without the dissolved-out interference amount fraction per unit time. It can thus be a stream of raw material of a specific size class or other property of the raw material used for fractionation, which is provided by the respective purification device after fractionation and purification. The throughput can be determined by means of a suitable measuring device, wherein in particular a gravimetric and / or a volumetric measuring device and / or radiation-related measuring device, in particular an X-ray device, can be used. The invention is moreover not restricted to which manipulated variables are changed in at least one of the process steps A), B), C), D) or E).Preferably, however, in process step A), a supply quantity or a variable of the raw material dependent thereon is changed as a first manipulated variable.The supply amount can be temporarily increased or decreased, thereby correspondingly increasing or decreasing the supplied amount of raw material per unit time. This has the effect that the amount of raw material located in the treatment plant is correspondingly changed, the amounts of material in the fractions are increased or decreased, and the feed and discharge rates to or from the cleaning devices can be adjusted by means of the algorithm in such a way that the fraction of interference amount can be reduced or increased as required, preferably limited by a maximum permissible fraction of interference material.Furthermore, the composition of the supply quantity can also be changed as a first manipulated variable. If the amount provided is in different qualities, wherein quality can describe, for example but not just the ratio of fine particles to coarse particles or solid wood particles to residual wood particles, these amounts provided can be provided in a mixing ratio which optimally loads the treatment plant.Additionally or alternatively, in process step B) a fraction characteristic variable, in particular a fraction distribution, fraction division and / or fraction division, is changed as a second manipulated variable of at least one fraction.The fraction distribution can describe the quantitative, spatial presence of raw material in the different fractions. The fraction distribution can be exemplified as a bar histogram, wherein the bars each represent a size class and the height of the bars can indicate the amount of raw material present therein. By changing the fraction distribution, the number of size classes in all or part of all present sizes of the parts of the raw material can be changed. Using the example of the bar histogram, one bar of the histogram would thus be divided into a plurality of bars or a plurality of bars would be combined to form one bar, as a result of which the heights of the respective bars can change. This therefore makes it possible to determine the size of a fraction and thus also the amount of raw material which is to be fed to a specific cleaning device.The fraction division can describe the subdivision of the present raw material sizes into different size classes. In this case, it therefore indicates which sizes are combined to form a size class. As illustrated by the bar histogram described above, the fraction split is determined by the width of the bars.The fractional division can describe the relative position of the present size size classes with respect to one another. In this case, the fractional division thus depends on the upper and lower limits of individual size classes. Using the example of the bar histogram, the fractional division is thus manifested in the position of the bars relative to one another.It is within the scope of the advantageous development that the fraction characteristic variable is not dependent on the sizes of the parts of the raw material, but on another criterion, for example a degree of moisture or another material-related characteristic variable.In process step C), a feed rate of at least one of the fractions into the cleaning devices can be changed as a third control variable.The feed rate may be regarded as a characteristic quantity indicating an amount per unit time at which the raw material of a fraction is fed to a purification device. In particular, it is within the scope of the advantageous development that a conveying section between a fractionating device and one of the cleaning devices comprises a material buffer to which fractionated material can be fed at a first rate and discharged at a second rate, wherein the first and the second rate can differ from one another. As a result, the feed rate into the cleaning device can be varied independently of the remaining feed rates in the plant.In process step D), a cleaning parameter of one of the cleaning devices can furthermore be changed as a fourth control variable.The cleaning parameter can be dependent on the type of cleaning device. If the cleaning device is based on a magnetic principle of action, the cleaning parameter can be a magnetic characteristic variable, in particular a magnetic field strength, or a characteristic variable dependent thereon. In the case of an inductive cleaning principle, the cleaning parameter can be an electrical characteristic variable, in particular a current intensity or voltage, or a characteristic variable dependent thereon. In a mechanical cleaning device, the cleaning parameter can be a mechanical characteristic variable, for example a force or a speed, or be dependent thereon. A pneumatic cleaning device can accordingly be a pneumatic characteristic variable, for example a pressure or air volume flow, or the characteristic variable can be dependent thereon. In a chemical cleaning device, the cleaning parameter may indicate a chemical aggressiveness or a process parameter, for example temperature or pressure. As already mentioned above, it is also within the scope of the invention that a sensor-based detection of the interfering substances takes place so that they can be removed from the load. In particular, an acoustic method, an optical method, preferably by means of color and / or shape recognition, a thermographic method, a radiographic method, in particular by means of X-ray radiation or an electromagnetic method, in particular by means of near-infrared radiation, can be used for this purpose.Before carrying out the process steps A), B), C), D), E), a reference raw material is preferably provided in the treatment plant and a material-related process parameter of the reference raw material is detected by means of a reference measurement system and evaluated by means of a reference algorithm. The algorithm is trained by means of the detection data of the reference measurement system and the evaluation data of the reference algorithm.The reference raw material can be a material whose properties are determined with high precision before carrying out the process steps A), B), C), D), E) in order to create a basis for comparison for training the algorithm. The reference measurement system can be identical to the measurement system used in process steps A), B), C), D), E) in order to determine a material-related process parameter. The reference algorithm does not necessarily also have to be based on the use of artificial intelligence. Instead, an analytical method can be used to determine the material-related process parameters and to output a measured variable. The data determined in this case can then be used to train the algorithm. In this case, the algorithm is therefore used not only for controlling the processing plant, but also for determining material-related process parameters.The method according to the invention or one of its advantageous developments for producing a workpiece is preferably carried out, wherein after method step E) the prepared starting material is processed to form the workpiece. In particular, before method step A), the raw material can be at least partially obtained by recycling.The object is also achieved by a processing plant having a providing means which is provided for conveying a raw material, in particular for producing a workpiece having a multiplicity of different constituents. The plant also comprises a fractionating device which is arranged to receive the raw material from the supplying means and which is adapted to divide the raw material into a plurality of fractions. In addition, the plant has a plurality of conveying means and cleaning devices, wherein the conveying means are provided for supplying the fractions to the cleaning devices, and wherein the cleaning devices are each provided for removing impurities from the fractions as impurity amount fraction. A discharging means is provided for receiving and discharging the processed fractions as a processed raw material from the cleaning apparatuses.It is essential that the treatment plant has at least one measuring device which is configured to record a material-related process parameter in the raw material received in the treatment plant. A control device is connected in terms of signal technology to the at least one measuring device, the provision means, the fractionating device, the conveying means and the cleaning devices and is configured to evaluate the material-related process parameter by means of an algorithm based on an artificial intelligence and to change at least one manipulated variable of the processing plant as a function of a processing target variable which is dependent at least on a throughput of the processed starting material and / or a cleaning efficiency of the processing plant.With regard to the treatment plant according to the invention, the statements regarding the method according to the invention and its advantageous developments apply accordingly. In particular, the processing plant is designed to carry out the method according to the invention or an advantageous development thereof.The processing plant is preferably part of a processing plant which is provided for producing a workpiece from the processed starting material. In particular, this can be a wood processing plant for producing plate-shaped wood workpieces, which preferably comprises at least one press for producing the plate-shaped wood workpiece.Some advantages of the invention are explained below with reference to an exemplary embodiment. The single FIG. 1 shows a schematic illustration of a processing plant for the processing of raw material for producing wood material boards.The processing plant 1 shown in FIG. 1 serves for processing a recycled raw material 2 for the production of a wood material board. The raw material 2 is present as bulk material and contains lignocellulosic material, in particular wood parts of different sizes and interfering substances in the form of metal (ferrous and non-ferrous metals), inert materials (glass, stones, ceramic), rubber, mineralics or plastic residues which were already present during the recycling of a raw material to a raw material 2 or were added to it. Before the raw material 2 can be used for the production of wood material boards, these contaminants should be removed from the raw material 2 to a predetermined degree so as not to adversely affect the product quality of the wood material board to be produced. In this case, it may be desirable to remove the impurities completely or only partially from the raw material 2.The processing plant 1 comprises a supply means 3 in the form of an endlessly revolving and driven conveyor belt, which is provided to receive the raw material 2 for the production of plate-shaped workpieces with a plurality of fractions and to supply it to the processing plant 1 via a supply means 3 in the form of a conveyor belt. Alternatively, this could also be a vibrating trough or comparable conveying meansFrom the supply means 3, the raw material 2 passes into a fractionating device 4, which in the embodiment shown in FIG. 1 comprises a roller bed 5, the rollers of which are arranged at increasing distances along the conveying axis. The portions of the raw material 2 on the roller bed 5 are either retained or fall between the rollers due to gravity, depending on the spacings of the rollers and the size of the portions. As a result, the raw material 2 is spatially sorted according to size. Below the roller bed 5 there are a plurality of funnel-like collecting devices 6, 7, 8, 9. By adjusting the separating walls of the collecting devices 6, 7, 8, 9, it is possible to set which orders of magnitude of the raw material 2 are to be picked up by the collecting devices 6, 7, 8, 9, respectively, as a result of which the raw material 2 is divided into a plurality of fractions 10, 11, 12, 13 and fed to the conveying means 14, 15, 16, 17. This can also be achieved by adjusting the conveying speed of the roller bed 5.The conveying means 14, 15, 16 serve to feed the fractions 10 and 11 and 12 to the cleaning devices 18, 19, 20. The conveying means 17, on the other hand, serves to feed the fraction 13, which substantially comprises oversizes and agglomerates, to a post-grinder 21.The fractions 10, 11, 12 are freed from impurities by means of the cleaning devices 18, 19, 20, which impurities are respectively removed as impurity quantity components 22, 23, 24 and are guided out of the treatment plant 1 via the removal conveyors 25, 26, 27. The processed fractions 10', 11', 12' are subsequently combined and provided as processed starting material 48 for further processing. Alternatively, the individual processed fractions 10', 11', 12' can also be supplied separately to further processing.A typical challenge in the operation of such treatment plants 1 is that the raw material 2 can be present in a very heterogeneous form and its properties can vary greatly depending on the origin. This can lead to the material distribution between the fractions 10, 11, 12 being present in an uneven manner and the cleaning devices 18, 19, 20 being correspondingly loaded in an uneven manner. In order to increase the plant efficiency of the processing plant 1, a control unit 28 is provided, which is configured to execute an algorithm based on an artificial intelligence and to control the processing plant 1.The control unit 28 has a plurality of signal inputs and a plurality of signal outputs. The control unit 28 is connected via the signal inputs to a plurality of measuring devices, with which various material-related process parameters are detected in the processing plant 1 and input into the control unit 28. The entered process parameters are evaluated by means of the artificial intelligence. Control signals are output via the signal outputs in order to control one or more of the above-described devices of the processing plant 1. With reference to FIG. 1, a difference between signal inputs and signal outputs is indicated with reference to arrow directions.The supply means 3 is equipped with an optical measuring device 29 which is designed to record the constituents of the raw material 2 and the size distribution thereof. These are input to the control unit 28 via the signal input 30. Additionally or alternatively, another measuring device 29 could also be used, for example on the basis of a near infrared method.The fractionating device 4 comprises in the present case two optical measuring devices 31, which serve to detect a quantity distribution of the raw material 2 on the roller bed 5 and to input it via the signal inputs 32 into the control unit 28.The conveying means 14, 15, 16, 17 each have a gravimetric measuring device, by means of which it is possible to detect the throughput of the fractions 10, 11, 12, 13 and input them into the control unit 28 via the signal inputs 33, 34, 35, 36.The discharge conveyors 25, 26, 27 likewise have a gravimetric measuring device, by means of which it is possible to determine the throughput of the discharged impurity quantity of the impurity quantity fractions 22, 23, 24 from the fractions 10, 11, 12 and to input it into the control unit 28 via the signal inputs 37, 38, 39.The discharged amount of processed raw material 48 is also detected by a measuring device and input to the control unit 28 via the signal input 40.Further measurements can also be made in the conveying devices 25, 26, 27. For example, the wood fraction can be measured in a pollutant fraction. In the conveying devices 10, 11, 12, the residual impurity content can also be determined. Information about the efficiency of the previously connected cleaning devices can be determined via this.As explained above, the artificial intelligence algorithm implemented on the control unit 28 is configured to evaluate the above-described signal inputs 30, 32, 33, 34, 35, 36, 37, 38, 39, 40 and to control one or more devices of the processing plant 1. It is particularly advantageous that the control 28 of the individual devices can be effected not only as a function of their own process parameters, but that the artificial intelligence is capable of regulating one of the devices across the installation as a function of the process parameter of another device. Moreover, it is decisive that the regulation of the treatment plant ( 1) can take place as a function of one or more predeterminable target variables, in particular a cleaning efficiency of the treatment plant ( 1) and a discharge throughput.The delivery rate of the provision means 3 can thus be changed as required via the control input 41.The control inputs 42, 43 and 44 serve to adjust a subdivision of the raw material 2 to be processed into the fractions 10, 11, 12, 13 by adjusting the separating walls of the elements 6, 7, 8, 9. As a result, the fractions 10, 11, 12 can be present in such a way that the cleaning devices 18, 19, 20 can be supplied with quantities of raw material (2) to be cleaned or fractions (10, 11, 12) to be cleaned that are adjusted according to requirements. The cleaning parameters of the cleaning devices 18, 19, 20 can be adjusted via the control inputs 45, 46, 47 in order to be able to set a desired degree of cleaning.List of reference characters1 Treatment plant 2 Raw material 3 Providing means 4 Fractionating device 5 Roller bed 6 Collecting device 7 Collecting device 8 Collecting device 9 Collecting device 10, 10' Fraction 11, 11' Fraction 12, 12' fraction 13 fraction 14 conveying means 15 conveying means 16 conveying means 17 conveying means 18 cleaning device 19 cleaning device 20 cleaning device 21 post-grinder 22 interference amount fraction 23 interference amount fraction 24 interference amount fraction 25 unloading conveyor 26 unloading conveyor 27 unloading conveyor 28 control unit 29 measuring device 30 signal input 31 measuring device 32 signal input 33 signal input 34 signal input 35 signal input 36 signal input 37 signal input 38 signal input 39 signal input 40 signal input 41 control input 42 control input 43 control input 44 control input 45 control input 46 control input 47 control input 48 starting material (processed) 49 application means

Claims

Method for processing a raw material (2) by means of a processing plant (1), comprising the process steps: A) providing raw material (2) in the processing plant (1), wherein the raw material (2) has a plurality of different constituents; B) dividing the raw material (2) into a plurality of fractions (10, 11, 12). C) feeding the fractions (10, 11, 12) into a plurality of cleaning devices (18, 19, 20); D) cleaning the fractions (10, 11, 12) in the cleaning devices (18, 19, 20), wherein in each case an amount of impurity is removed from the fractions (10, 11, 12) as an impurity amount fraction (22, 23, 24); E) Application of the fractions (10', 11', 12') as processed starting material (48); characterized in that at least one material-related process parameter is detected at least in one of the process steps A), B), C), D) or E) and evaluated by means of an algorithm on the basis of an artificial intelligence, and in that at least one manipulated variable of one of the process steps A), B), C), D) or E) is changed as a function of a processing target variable which is dependent at least on a throughput of the processed starting material (48) and / or a cleaning efficiency of the processing plant (1).Method according to Claim 1, in which in process step A) at least one material property of the raw material (2), in particular an optical feature, preferably a ratio between a fines fraction and a coarse fraction and / or a ratio between a solid wood fraction and a residual wood fraction, and / or a moisture, is detected as a first material-related process parameter.Method according to Claim 1 or 2, in which in process step B) at least one optical feature of at least one of the fractions (10, 11, 12), in particular a quantity distribution of the fractionated raw material (2), is detected as a second process parameter.Method according to one of the preceding claims, in which at least one throughput of the fractions (10, 11, 12) is detected as a third process parameter in process step C).Method according to one of the preceding claims, in which in process step D) at least one unfrired interference quantity fraction (22, 23, 24) or a purity level of at least one of the fractions (10, 11, 12) is detected as a fourth process parameter.Method according to one of the preceding claims, in which in process step E) at least one throughput of the processed starting material (48) of a fraction (10', 11', 12') is detected as a fifth process parameter.Method according to one of the preceding claims, in which in process step A) a supply quantity or a variable of the raw material (2) dependent thereon is changed as a first control variable.Method according to one of the preceding claims, in which in process step B) a fraction characteristic variable, in particular a fraction distribution, fraction division and / or fraction division, is changed as a second manipulated variable of at least one fraction (10, 11, 12).Method according to one of the preceding claims, in which in process step C) a feed rate of at least one of the fractions (10, 11, 12) into the cleaning devices (18, 19, 20) is changed as a third control variable.Method according to one of the preceding claims, in which in process step D) a cleaning parameter of one of the cleaning devices (18, 19, 20) is changed as a fourth control variable.Method according to one of the preceding claims, in which, before the process steps A), B), C), D), E) are carried out, a reference raw material is provided in the treatment plant (1) and a material-related process parameter of the reference raw material is detected by means of a reference measurement system and evaluated by means of a reference algorithm, and the algorithm is trained by means of the detection data of the reference measurement system and the evaluation data of the reference algorithm.Method for producing a workpiece, comprising a method for processing a raw material (2) according to one of Claims 1 to 11, wherein before method step A) the raw material (2) is preferably at least partially obtained by recycling; and after method step E) the processed raw material (48) is processed to form the workpiece.Treatment plant (1), in particular for carrying out a method according to one of Claims 1 to 11, having a provision means (3) which is provided for conveying a raw material (2) having a multiplicity of different constituents, a fractionating device (4) which is arranged in such a way as to pick up the raw material (2) from the provision means (3) and which is designed to divide the raw material (2) into a multiplicity of fractions (10, 11, 12), a multiplicity of conveying means (14, 15, 16, 17) and cleaning devices (18, 19, 20), wherein the conveying means (14, 15, 16) are provided for supplying the fractions (10, 11, 12) to the cleaning devices (18, 19, 20), and wherein the cleaning devices (18, 19, 20) are each provided for removing interfering substances from the fractions (10, 11, 12) as interfering substance quantity fraction (22, 23, 24), and having a discharge means (49), which is provided for receiving and discharging the processed fractions (10, 11, 12) as processed starting material (48) from the cleaning devices (18, 19, 20), characterized at least one measuring device (29, 31) which is configured to record a material-related process parameter in the raw material (2) received in the processing plant (1) and by a control device (28) which is connected in terms of signal technology to the measuring device (29, 31), the provision means (3), the fractionating device (4), the conveying means (14, 15, 16, 17) and the cleaning devices (18, 19, 20) and is configured to evaluate the material-related process parameter by means of an algorithm based on an artificial intelligence and to change at least one control variable of the processing plant (1) as a function of a processing target variable which is dependent at least on a throughput of the processed raw material (48) and / or a cleaning efficiency of the processing plant (1).Processing plant for producing a workpiece with a processing plant according to claim 13.

Images