Method for producing material boards in a production plant, production plant, computer-program product and use of a computer-program product
A method for producing material boards in a production plant in which apparatuses form a material into a mat that is pressed to obtain the material board which has specific quality parameters. The production plant and/or the apparatuses are controlled in an open-or closed-loop manner by a controller, which preferably includes a programmable logic controller, and input parameters are received, processed and/or output by the controller. The input parameters are formed at least from settable product parameters for the material board to be produced, from settable and/or recorded plant parameters of the production plant and/or the apparatuses and/or from recorded material parameters. A quality value of at least one quality parameter of the material board to be produced is determined based on the input parameters by an algorithm based on artificial intelligence. The algorithm is trained or formed by a database which has at least one quality parameter and input parameters correlating with the quality parameter.
1 . A method for producing a material board ( 12 ) in a production installation ( 1 ), comprising:
forming, by a first subset of apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ) of a production installation ( 1 ) based on a first subset of formed input parameters ( 23 ) and ascertained quality values ( 24 ), a material ( 10 ) into a mat ( 11 ) and pressing, by a second subset of the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ) of the production installation ( 1 ) based on a second subset of the formed input parameters ( 23 ) and the ascertained quality values ( 24 ), the mat ( 11 ) to produce a material board ( 12 ), wherein the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ) of the production installation ( 1 ) comprise one or more of
temporary stores of the material ( 10 ),
a drying stage,
a spreading apparatus ( 2 ) comprising spreading heads,
a gluing apparatus ( 3 ),
a pressing apparatus ( 4 ) comprising cylinders and heating plates,
a comminuting apparatus ( 5 ),
a drying apparatus ( 6 ),
a sorting apparatus ( 7 ),
a pre-pressing apparatus ( 8 ), and
a separating apparatus ( 9 );
measuring, by a first subset of measuring apparatuses of the production installation ( 1 ), first actual values of material parameters of the material ( 10 );
measuring, by a second subset of measuring apparatuses of the production installation ( 1 ), second actual values of installation parameters of the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ) and the production installation ( 1 );
acquiring, by an interface ( 25 ) of a controller ( 20 ) that is directly connected to each of the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ) and to each of the measuring apparatuses, the measured first actual values of the material parameters of the material ( 10 ) and the measured second actual values of the installation parameters of the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ) and the production installation ( 1 ),
acquiring, by the interface ( 25 ) of the controller ( 20 ), first setpoint values of the material parameters of the material ( 10 ), second setpoint values of the installation parameters of the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ) and the production installation ( 1 ), and third setpoint values of product parameters of the material board ( 12 ) to be produced;
forming, by the controller ( 20 ), input parameters ( 23 ) based on the acquired first setpoint values, the acquired second setpoint values, the acquired third setpoint values, the measured first actual values, and the measured second actual values, wherein each of the formed input parameters ( 23 ) comprises at least one of a respective time value or a respective position value;
ascertaining, by the controller ( 20 ) using an artificial intelligence-based algorithm ( 32 ) and based on the formed input parameters ( 23 ), the quality values ( 24 ) of quality parameters of the material board ( 12 ), the algorithm ( 32 ) having been trained, or formed, by a database ( 30 ) stored in a memory ( 27 ) of the controller ( 20 ) and comprising at least one stored quality parameter and stored input parameters correlating with the at least one stored quality parameter; and
transmitting, by the interface ( 25 ) of the controller ( 20 ), the formed input parameters ( 23 ) and the ascertained quality values ( 24 ) to the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ), controlling or regulating the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ) and producing the material board ( 12 ) based on the transmitted formed input parameters ( 23 ) and the ascertained quality values ( 24 ).
2 . The method as claimed in claim 1 , further comprising at least one of normalizing, by the controller ( 20 ), or aggregating, by the controller ( 20 ), the input parameters ( 23 ).
3 . The method as claimed in claim 2 , wherein at least one of (a) the normalizing of the input parameters ( 23 ) includes normalizing, by the controller ( 20 ), the input parameters over at least one of time or a length of at least one of the production installation ( 1 ) or one or more of the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ), or (b) the aggregating of the input parameters ( 23 ) includes forming, by the controller ( 20 ), clusters within one or more of the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ).
4 . The method as claimed in claim 1 , wherein the database ( 30 ) is formed at least from a dataset formed from data relating to the at least one stored quality parameter and the stored input parameters correlating with the at least one stored quality parameter, and the database includes at least an additional dataset, and the dataset differs from the additional dataset of the database ( 30 ) in at least one of the quality parameter or the input parameter correlating with the quality parameter.
5 . The method as claimed in claim 4 , wherein the data or the dataset are at least one of (a) normalized, by the controller ( 20 ), over at least one of time or to at least one of a length of the production installation ( 1 ) or one or more of the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ) or (b) aggregated, by the controller ( 20 ), by forming data clusters.
6 . The method as claimed in claim 1 , further comprising checking, by the controller ( 20 ), the artificial intelligence-based algorithm ( 32 ) using a test database ( 33 ) stored in the memory ( 27 ), wherein the test database ( 33 ) includes test datasets containing test quality parameters and test input parameters correlating with the test quality parameters, and wherein the test datasets differ from datasets of the database ( 30 ).
7 . The method as claimed in claim 1 , further comprising first checking, by the controller ( 20 ) using the artificial intelligence-based algorithm ( 32 ), an alteration in the input parameters ( 23 ) for an effect on the ascertained quality value ( 24 ).
8 . The method as claimed in claim 1 , further comprising optimizing, by an optimization computer ( 22 ) of the controller ( 20 ) using at least one of the artificial intelligence-based algorithm ( 32 ) or at least one other artificial intelligence-based algorithm ( 34 ), the ascertained quality values ( 24 ), the at least one other artificial intelligence-based algorithm ( 34 ) having been formed by training using at least one of the database ( 30 ) or another database.
9 . The method as claimed in claim 8 , wherein the optimizing the ascertained quality values ( 24 ) comprises visualizing altered input parameters ( 28 ).
10 . The method as claimed in claim 8 , wherein the optimizing the ascertained quality values ( 24 ) comprises optimizing the ascertained quality values ( 24 ) at least one of before production, at a start of production, during production, or after a predetermined time interval.
11 . The method as claimed in claim 1 , further comprising at least one of outputting, by the interface ( 25 ) to a display device of the production installation ( 1 ), a warning or switching on, by the controller ( 20 ), an optimization computer ( 22 ) upon a difference between the ascertained quality value ( 24 ) and a predefined setpoint value or a setpoint value range; or outputting, by the interface ( 25 ) to a display device of the production installation ( 1 ), a message if adherence to the quality value ( 24 ) in the setpoint value range is achieved based on altered input parameters ( 28 ), and displaying the altered input parameters ( 28 ) to a user.
12 . The method as claimed in claim 1 , wherein the algorithm ( 32 ) is based on or includes at least one method or a modification thereof from one of the following groups:
linear regression, polynomial regression, functional regression, K nearest neighbors regression, random forest regression, support vector regression, neural networks, recurrent neural networks, convolutional neural networks, residual networks, Bayesian networks, K nearest neighbors classification, decision trees, random forests, naive Bayes, or support vector machines.
13 . The method as claimed in claim 1 , wherein a physical model is incorporated into the algorithm ( 32 ) to ascertain the quality value ( 24 ).
14 . The method as claimed in claim 1 , wherein the algorithm includes multiple algorithms ( 32 , 34 ) each configured to independently ascertain a predicted value for the quality value ( 24 ) for the at least one quality parameter.
15 . The method as claimed in claim 14 , wherein the predicted values independently ascertained by the multiple algorithms ( 32 , 34 ) are at least one of offset against one another or compared, and a combined result is used as the quality value ( 24 ) for a basis for optimization.
16 . The method as claimed in claim 1 , further comprising at least one of (a) deriving, by the controller ( 20 ) using the artificial intelligence-based algorithm ( 32 ) and based on the formed input parameters ( 23 ), information relating to at least one of wear or abnormal behavior or (b) forecasting, by the controller ( 20 ) using the artificial intelligence-based algorithm ( 32 ) and based on the formed input parameters ( 23 ), information relating to at least one of wear or abnormal behavior for at least one of state monitoring, predictive maintenance detection, or failure detection.
17 . The method as claimed in claim 1 , further comprising checking, by the controller ( 20 ), the artificial intelligence-based algorithm ( 32 ) at least one of before or during ongoing production by comparing one of the ascertained quality values ( 24 ) with a measured quality value of the quality parameter of the material board ( 12 ), and if a difference between the one of the ascertained quality values ( 24 ) and the measured quality value is detected that is above a stipulated threshold value, retraining, by the controller ( 20 ), the artificial intelligence-based algorithm ( 32 ).
18 . The method as claimed in claim 1 , further comprising at least one of transmitting, by the interface ( 25 ), or receiving, by the interface ( 25 ), at least one of the artificial intelligence-based algorithm ( 32 ), at least one dataset of the production installation ( 1 ), or the database ( 30 ).
19 . The method as claimed in claim 1 , wherein the material parameters of the material ( 10 ) comprise one or more of
a type of the material ( 10 ),
a density of the material ( 10 ),
a moistness of the material ( 10 ),
a composition of the material ( 10 ),
a respective temperature of the material ( 10 ) in each of the temporary stores of the material ( 10 ),
a respective temperature of the material ( 10 ) in each of the temporary stores of the material ( 10 ),
a type and amount of binder used by the gluing apparatus ( 3 ),
an amount of binder used by the gluing apparatus ( 3 ), and
a temperature of the material ( 10 ) before entering the pressing apparatus ( 4 ).
20 . The method as claimed in claim 1 , wherein the installation parameters of the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ) and the production installation ( 1 ) comprise one or more of
a respective fill level of each of the temporary stores of the material ( 10 ),
a temperature of the drying stage for drying the material ( 10 ) before the forming of the material ( 10 ) into the mat ( 11 ),
a use and a type of preliminary heating for the mat ( 11 ),
a spreading width in which the mat ( 11 ) is spread by the spreading apparatus ( 2 ),
a number of the spreading heads of the spreading apparatus ( 2 ),
a press length of the pressing apparatus ( 4 ),
a respective pressing pressure of each of the cylinders of the pressing apparatus ( 4 ),
a respective temperature of each of the heating plates of the pressing apparatus ( 4 ),
respective power consumption values for each of the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ) and for the production installation ( 1 ),
a power consumption value for the production installation ( 1 ),
a room temperature of the production installation ( 1 ), and
an air humidity of the production installation ( 1 ).
21 . The method as claimed in claim 1 , wherein the product parameters of the material board ( 12 ) to be produced comprise one or more of
a type of the material board ( 12 ) to be produced,
a thickness of the material board ( 12 ) to be produced, and
a strength of the material board ( 12 ) to be produced.
22 . The method as claimed in claim 1 , wherein the ascertained quality values ( 24 ) of the quality parameters of the material board ( 12 ) comprise one or more of
a first strength value for a major axis of the material board ( 12 ),
a second strength value for a minor axis of the material board ( 12 ),
a first transverse tensile strength value for a major axis of the material board ( 12 ),
a second transverse tensile strength value for a minor axis of the material board ( 12 ),
a first modulus of elasticity value for a major axis of the material board ( 12 ),
a second modulus of elasticity value for a minor axis of the material board ( 12 ), and
a surface quality value of the material board ( 12 ).
23 . The method as claimed in claim 1 , the forming the material ( 10 ) into the mat ( 11 ) and pressing the mat ( 11 ) to produce the material board ( 12 ) comprising:
forming, by the first subset of apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ) further based on altered input parameters ( 28 ), the material ( 10 ) into the mat ( 11 ) and pressing, by the second subset of the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ) further based on the altered input parameters ( 28 ), the mat ( 11 ) to produce the material board ( 12 ), and
the method further comprising:
altering, by the controller ( 20 ) using the artificial intelligence-based algorithm ( 32 ) and based on the ascertained quality values ( 24 ), one or more of the formed input parameters ( 23 ) to form altered input parameters ( 28 ); and
transmitting, by the interface ( 25 ) of the controller ( 20 ), the altered input parameters ( 28 ) to the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ) for controlling or regulating the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ) to produce the material board ( 12 ) further based on the altered input parameters ( 28 ).
24 . The method as claimed in claim 1 ,
the forming the input parameters ( 23 ) comprising forming, by a programmable logic controller ( 21 ) of the controller ( 20 ), the input parameters ( 23 ) based on the acquired first setpoint values; and
the ascertaining the quality values ( 24 ) comprising ascertaining, by the programmable logic controller ( 21 ) of the controller ( 20 ) using the artificial intelligence-based algorithm ( 32 ) and based on the formed input parameters ( 23 ), the quality values ( 24 ) of the quality parameters of the material board ( 12 ).
25 . A production installation ( 1 ) for producing material boards ( 12 ) that have specific quality parameters, the production installation ( 1 ) comprises:
apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ) configured to
glue, by a first subset of the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ) based on a first subset of formed input parameters ( 23 ) and ascertained quality values ( 24 ), a material ( 10 ),
form, by a second subset of the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ) based on a second subset of the formed input parameters ( 23 ) and the ascertained quality values ( 24 ), a mat ( 11 ) from the glued material ( 10 ), and
press, by a third subset of the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ) based on a third subset of the formed input parameters ( 23 ) and the ascertained quality values ( 24 ), the mat ( 11 ) to produce a material board ( 12 ),
the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ) comprising one or more of
temporary stores of the material ( 10 ),
a drying stage,
a spreading apparatus ( 2 ) comprising spreading heads,
a gluing apparatus ( 3 ),
a pressing apparatus ( 4 ) comprising cylinders and heating plates,
a comminuting apparatus ( 5 ),
a drying apparatus ( 6 ),
a sorting apparatus ( 7 ),
a pre-pressing apparatus ( 8 ), and
a separating apparatus ( 9 );
measuring apparatuses configured to
measure, by a first subset of the measuring apparatuses, first actual values of material parameters of the material ( 10 ), and
measure, by a second subset of the measuring apparatuses, second actual values of installation parameters of the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ) and the production installation ( 1 ); and
a controller ( 20 ) that is directly connected to each of the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ) and to each of the measuring apparatuses, the controller ( 20 ) comprising an interface ( 25 ) and a memory ( 27 ), the controller ( 20 ) configured to
acquire, by the interface ( 25 ), the measured first actual values of the material parameters of the material ( 10 ) and the measured second actual values of the installation parameters of the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ) and the production installation ( 1 ),
acquire, by the interface ( 25 ), first setpoint values of the material parameters of the material ( 10 ), second setpoint values of the installation parameters of the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ) and the production installation ( 1 ), and third setpoint values of product parameters of the material board ( 12 ) to be produced,
form input parameters ( 23 ) based on the acquired first setpoint values, the acquired second setpoint values, the acquired third setpoint values, the measured first actual values, and the measured second actual values,
each of the formed input parameters ( 23 ) having at least one of an attributed time value or position value,
ascertain, using an artificial intelligence-based algorithm ( 32 ) and based on the formed input parameters ( 23 ), the quality values ( 24 ) relating to quality parameters of the material board ( 12 ), the algorithm ( 32 ) being trainable, or formable, by a database ( 30 ) stored in the memory ( 27 ) and comprising at least one stored quality parameter and stored input parameters correlating with the at least one stored quality parameter, and
transmit, by the interface ( 25 ), the formed input parameters ( 23 ) and the ascertained quality values ( 24 ) to the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ), controlling or regulating the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ) and producing the material board ( 12 ) based on the transmitted formed input parameters ( 23 ) and the ascertained quality values ( 24 ).
26 . The production installation ( 1 ) as claimed in claim 25 , the controller ( 20 ) further configured to at least one of normalize or aggregate the input parameters ( 23 ) over at least one of time or a length of at least one of the production installation ( 1 ) or one of the apparatuses ( 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 ).