Interface component, network, method for monitoring the state of at least one distributed component, and computer program

Abstract

The invention relates to an interface component (10) of a distributed component (20a, 20b, 20c) of a network, wherein the network has a plurality of distributed components (20a, 20b, 20c) each with an interface component (10), wherein the interface component (10) has the following features: a processor (16) which is designed to determine a state (16z) of a decentral process model (16p) belonging to the respective distributed component (20a, 20b, 20c); a communication interface (12) which is designed to exchange external data (12d) for the process model (16p) belonging to a further distributed component (20a, 20b, 20c) of the plurality of distributed components (20a, 20b, 20c); an internal interface (14) which is designed to receive component parameters for the process model (16p) belonging to the distributed component (20a, 20b, 20c); wherein the state (16z) is determined based on the component parameters and / or the incoming external data (12d).

Description

[0001] INTERFACE COMPONENT, NETWORK, METHOD FOR CONDITION MONITORING OF AT LEAST ONE DISTRIBUTED COMPONENT, AND COMPUTER PROGRAM

[0002] Description

[0003] Embodiments of the present invention relate to an interface component and to a network comprising a plurality of distributed components, each with an interface component. Further embodiments relate to a production plant or material sorting plant. Another embodiment relates to a method and a corresponding computer program.

[0004] In general, the invention lies in the field of distributed systems for optimizing physical process chains across processes, in particular processes across company and industry boundaries.

[0005] Process chains in industry comprise numerous process steps in a complex network that spans company boundaries. Each process step has numerous adjustment options that influence the outcome in terms of quality / quantity, energy costs, and CO2 emissions. Optimizing this system according to local (e.g., profit) or global cost functions (e.g., CO2 emissions) is desirable, but if this were performed centrally, all actors would have to disclose their trade secrets.

[0006] Therefore, there is a need for an improved approach.

[0007] The object of the present invention is to create a concept that enables the optimization of process chains across processes, companies and industry boundaries.

[0008] The problem is solved by the subject matter of the independent patent claims.

[0009] Embodiments of the present invention provide an interface component of a distributed component of a network. The distributed component can be, for example, a system in a process chain, e.g., a sorting system or a production system. The network comprises a plurality of distributed components, each with an interface component. The plurality of distributed components form, for example, the components of the process chain. Each interface component has a processor, a communication interface, and an internal interface. The processor is designed to determine a state of a decentralized process model associated with the respective distributed component. The communication interface is designed to exchange external data for the process model associated with another distributed component of the plurality of distributed components.For example, the exchange can occur with other interface components belonging to the other distributed components. The internal interface is designed to receive component parameters, such as sensor parameters for the process model belonging to the distributed component. The state is determined based on the component parameters and / or the incoming external data.

[0010] According to one embodiment, each decentralized process model has a state that is differentiable with respect to all component parameters and / or all incoming external data. According to embodiments, differentiable means that derivatives of the state are available according to component parameters, the model, and the communication with neighboring process models. According to further embodiments, the state of the respective component, together with the decentralized process model, forms a digital twin of the component. If, according to embodiments, it is assumed that a decentralized process model including state is also provided in the respective interface components of the further distributed components or neighboring distributed components, then - in other words - each or at least some interface components can be assigned to the distributed components orneighboring components have a digital twin. Communication between the neighboring digital twins takes place via the communication interface.

[0011] Embodiments of the present invention are based on the finding that the use of interface components for distributed components of a system enables machine learning across the process chain, thus enabling global optimization to be achieved if the distributed components in a process chain are expanded by corresponding interface components. The interface components can be used to map the setting options as well as different qualities and quantities for each process step. Higher-level parameters, such as energy costs or CO2 emissions, can also be mapped for each process step in this way, specifically in a differentiated manner, i.e., in relation to the incoming data, i.e., the parameters from the neighboring distributed components and to the internal process data.According to embodiments, the communication interface is configured to receive a higher-level parameter and / or a cost function associated with the higher-level parameter, such as a CCh target, and to perform the optimization. According to embodiments, the optimization is based on component parameters and / or incoming external data (information regarding neighboring components and / or higher-level targets).

[0012] By using the interface component, it is possible for each distributed component coupled to the interface component to maintain a decentralized process model, which, together with its current state, represents a digital twin of the component. This model can be used for modeling and optimization. Because the interface components exchange data (referred to above as external data) in a process chain across processes, companies, and industries via the communication interface, the dependency of the distributed component—modeled as a digital twin of the distributed component—is taken into account.

[0013] The interface component for the distributed component of a system enables machine learning and thus an optimizable overall system without all actors having to disclose their trade secrets.

[0014] According to embodiments, during optimization, only or partially or preferably gradients of component parameters and / or derivatives of component parameters and / or a corresponding cost function are exchanged as external data. Thus, it is advantageously not even necessary to transmit the parameter relevant for the one distributed component in full, since the change in this regard alone is sufficient through the exchange of gradients. Alternatively, a cost function, such as a global cost function, can also be transmitted in order to map an optimization goal to one or more distributed components. According to embodiments, the component parameters and / or process parameters are backpropagated as outgoing external data. According to embodiments, this can be the absolute value or also a gradient or a derivative.

[0015] According to further embodiments, in addition to gradients, absolute values ​​are also transmitted during state determination. Both gradients and absolute values ​​always include quantities that actually describe the material / energy / CO2 flow between the components of the distributed system. No internal information (internal state of the component, previous process steps, etc.) is exchanged.

[0016] According to embodiments, the processor of each interface component is designed to perform machine learning / optimization in a decentralized manner, e.g., taking into account the cost function explained above. Because the interface components are networked along the process flow / the real material flow in accordance with embodiments, an optimization of the overall system is enabled using the respective digital twins. In other words, in order to implement such a machine learning system for real material flows, the components of the overall system are each mapped as a digital twin and optimized in a decentralized manner. The digital twin consists of a differentiable process model in an associated state. The state of the respective digital twin is (constantly) decentralized based on component parameters, such as, for example,Sensor data, and estimated and updated through communication with neighboring components.

[0017] According to embodiments, the processor is configured to perform a local optimization with respect to the distributed component or a portion of the global optimization with respect to the entire system. According to further embodiments, the processor is configured to determine an optimization step size for optimizing the decentralized process model based on gradients and / or information about an uncertainty of the gradients.

[0018] According to preferred embodiments, the state is derivable with respect to all component parameters, communication with neighboring components, and the digital twin model. According to embodiments, a derivable state with respect to the decentralized model can mean that the decentralized model is adapted or updated according to a variant, e.g., because the situation, such as the boundary conditions, of the decentralized component has changed. A further conclusion from the derivability of the decentralized model is that there is a maintenance requirement for the distributed component. In this respect, information regarding a maintenance requirement can be determined based on the derivability with respect to the decentralized model. According to embodiments, this determination is performed by the processor.The reason for this is that by using the current component parameters and external data, it can be determined that the existing process model no longer fits, resulting in a mismatch between the distributed component and the digital twin. This mismatch can be resolved either by updating the decentralized model, i.e., the digital twin, or by returning the distributed component to the modeled state, e.g., through maintenance.

[0019] According to further embodiments, the decentralized model can be derived with respect to incoming external data. This means that, according to preferred embodiments, it is possible to identify which external variable has what influence on the decentralized model and, in particular, on the state of the component. In this respect, according to embodiments, the processor is configured to provide a derivation of the state based on the component parameters and / or the incoming external data and / or the process model. According to embodiments, the processor uses inferential statistics, such as Bayesian methods, to determine the state.

[0020] At this point, it should be noted that the interface components of the network are interconnected according to the material flow between multiple distributed components and the distributed component. This means that, according to exemplary embodiments, the state is determined based on incoming external data from further distributed components, taking into account the material flow between the multiple distributed components and the distributed component. In other words, this means that, in particular, the incoming external data from neighboring distributed components is taken into account.

[0021] Another embodiment provides a network comprising a plurality of distributed components, each with an interface component. The interface component is configured as explained above.

[0022] A further embodiment relates to a production plant or material sorting plant comprising a corresponding network.

[0023] A further embodiment provides a method comprising the following steps: determining a state of a decentralized process model associated with the respective distributed component; exchanging external data for the process model associated with another distributed component; obtaining a component parameter for the process model associated with the distributed component; wherein the state is determined based on the component parameters and / or the incoming external data.

[0024] According to embodiments, the method may include the step of mapping a digital twin associated with the distributed components. According to a further embodiment, it would be conceivable for the method to include the step of optimizing the decentralized process model.

[0025] According to embodiments, the method can be computer-implemented.

[0026] Embodiments of the present invention are explained with reference to the accompanying drawings. They show:

[0027] Fig. 1 is a schematic representation of an interface component associated with a distributed component according to a basic embodiment; and

[0028] Fig. 2 is a schematic representation of an exemplary application, here a material sorting system with several distributed components and interface components according to embodiments.

[0029] Before exemplary embodiments of the present invention are explained below with reference to the accompanying drawings, it should be noted that elements and structures with the same function are provided with the same reference numerals, so that the description of them is applicable to one another or interchangeable.

[0030] Fig. 1 shows an interface component 10 of a distributed component 20a of a network 20 of distributed components 20a-c (see Fig. 2). Each interface component 10 has the core components of a communication interface 12, an internal interface 14, and a processor 16.

[0031] As can be seen from Fig. 2, the distributed components 20a-c of the network can be a production plant or a material sorting plant. Here, a material sorting plant with a material flow 22 and three different plant components is shown as an example. For example, the first component 20a can be a conveyor device with material analysis, component 20b a material sorter, and component 20c a storage device. Each component 20a, 20b, and 20c is coupled to its own interface component 10a, 10b, and 10c, which can, on the one hand, receive component parameters or process parameters from components 20a, 20b, and 20c and, on the other hand, can communicate with the other interface components 10a / 10b / 10c.

[0032] For this purpose, each interface component has the communication interface 12 and the internal interface 14.

[0033] The internal interface 14 is used to exchange component parameters, such as sensor data or process data 12d, with the corresponding distributed components 20a, 20b, and 20c. For example, a sensor can be directly connected via the internal interface, via which data from components 20a, 20b, and 20c or control information, such as a set speed for the roller conveyor or a threshold value for the sorter, can be read into the interface components 10a, 10b, or 10c.

[0034] By means of the communication interface 14, data is exchanged with external components, i.e. with other distributed components, for example. For example, the interface component 10b can receive information about the material flow rate from the interface component 10a, on the basis of which separate process parameters are then set for the component 20b. Depending on the application or current objective, this information can be received either as an absolute value or as a relative value in the sense of faster than before or slower than before (possibly with associated change information or gradients). For example, when exchanging information for state assessment, absolute values ​​are exchanged, while during optimization it is primarily relative gradients / requirements that are used. In both cases, the relative or absolute values ​​relate to the same physical quantities, which in turn primarily relate to the material / energy / CO2 / etc.Describe the flow between the components. According to embodiments, the external data to be exchanged can include information regarding material, energy, CO2, etc. flow between the components.

[0035] The processor 16 uses a decentralized process model 16p to determine a state of the associated distributed component, i.e., component 20a for interface component 10a, or 20b for interface component 10b, or 20c for interface component 10c. The determination is made using the so-called decentralized twin, which represents the combination of the process model 16p and the state information 16z. Since there is always a dependency between the components 20a, 20b, and 20c, the digital twin is adapted to exchange the external data 12d.

[0036] The basis for the implementation is the above-described interface component 10 for the distributed components 20a, 20b, and 20c of the system 20. These interface components 10 enable machine learning for each component 10a, 10b, and 10c, thus creating an optimizable system. To implement such a machine learning system for real material flows, the individual components 20a, 20b, and 20c of the overall system are mapped as a digital twin. This digital twin consists of a differentiable process model 16p and an associated state 16z. The state 16z of the digital twin is continuously estimated and updated in a decentralized manner based on sensor data or through communication with neighboring components (modeling for the purpose of state determination). This state can be derived with respect to all component parameters, the communication with neighboring nodes, and the digital twin model.

[0037] According to exemplary embodiments, the optimization of the digital twin can be carried out in detail as follows: The interface components 10 for the distributed components 20 form the basis for optimizing any network. The derivative of a cost function is, for example, propagated backward through the network and multiplied at each node by the derivative of the node. This calculates and implements the change steps for each process parameter. After several steps, the system approaches the optimum.

[0038] In order to transfer this approach to machines that operate on a real material flow and therefore do not implicitly have derivatives, the digital twin is used for each process step. The digital twin consists of a differentiable model and a continuous state description (real number vector), and can thus assume (provide) the derivatives, i.e. the influence of a change in the input variables or parameter variables, for each state. The state of each digital twin is estimated as accurately as possible using local sensor data, the model and communication with other digital twins (decentralized data fusion). This is done, for example, using inferential statistics (Bayesian inference). This estimation result (e.g. maximum a posteriori estimate) must be differentiable with respect to all input variables (parameters, communication with other digital twins, models). This means:All models must also be differentiable. In principle, white-box, grey-box, and black-box models can be used.

[0039] This approach advantageously enables a digital twin to be generated / generated for any component, which models the dependency on the internal data (data directly concerning the respective component) and the external data (data indirectly concerning the respective component). Typically, the external data 12d that are directly adjacent to a component 20c in terms of material flow are particularly relevant, i.e., data from the external component 20b. This data can be determined from the digital twin of the interface component 10b, either based on the component parameters of component 20b or from other external data 12d. Each digital twin represents the respective component 20a, 20b, and 20c in their relevant dependencies, without the direct need to model the entire system. Conversely, however, the majority of the interface components 10a-10c model the entire system.This approach has significant advantages:.

[0040] Through the distributed system 10a, 10b, and 10c for optimizing process chains (see reference numeral 20) across processes, companies, and industries, global optimization can be achieved, on the one hand, by optimizing each process model 10p locally / decentrally and networking it with neighboring interface components or all other interface components or globally. On the other hand, the sensitivity of the individual data of the network participants can be maintained, since only the relevant data 12d is exchanged, or even only gradients are exchanged. Optimization here means adjusting the parameters of all processes affecting the material flow in such a way that a cost function is minimized. A cost function can be a societal goal such as CO2 reduction, or a company- or sector-related one, such as the scrap rate or negative profit.

[0041] Decentralized optimization also has positive effects on the cross-company approach: Using the interface component 10a / 10b / 10c for distributed components 20a / 20b / 20c of a machine learning system, process chains across different companies and sectors can be digitized and optimized. For this purpose, each process step or component 20a, 20b, 20c is provided with an interface 10a, 10b, 10c for backpropagation of gradients of the cost function, enabling optimization based on backpropagation in machine learning. A cost function as an external specification, e.g., from a legislator or another higher-level organization, means that not only process data from other components is exchanged as external data, but also complex functions that define a target, such as maximum CO2 emissions.This is actually only possible in the case of fully digital process steps, which is why the interface component 10 for distributed components 20 of a machine learning system was not directly applicable to physical process chains. To make this possible, the interface component 10 now supplements each process step or component 20 with a digital twin, which consists of a process model 16p and an associated state description 16z. The scope of such a digital twin is freely selectable by the responsible market player and can encompass a single process step, machine, component, or even the entire company. The interface component 10 explained above is then attached to the digital twin, thus enabling the backpropagation of gradients and thus the decentralized optimization of the process chain.To enable precisely this backpropagation, the state 16z of the digital twin is estimated as accurately as possible. This is estimated partly from local sensor data (see interface 14), but also relies on information 12d exchanged between the digital twins of neighboring process steps (see interface 12).

[0042] According to embodiments, the interface 10 can be extended to enable the exchange of additional information with associated uncertainty measures instead of just gradients (changes in a process parameter exchanged as external data 12d).

[0043] The implementation of this state estimation 16z can be carried out using inferential statistical methods (e.g., Bayesian inference, decentralized Kalman filtering, etc.), as shown in the examples. It is important that the implementation is fully differentiable, i.e., the derivatives of the state are available based on component parameters, the model 16p, and communication with neighboring digital twins. This serves the following purpose:

[0044] To optimize the process / component parameters, the derivative of these is known. • If machine 20 becomes worn, the associated model 16p can be adjusted / retrained.

[0045] • The communication between neighboring digital twins represents the physical material flow. For the global optimization of the process chain, the derivatives of the communicated information can therefore also be known. The uncertainty considerations in the communication between digital twins and in the state estimation are also crucial for the backpropagation of the gradients and thus the optimization, as this can also provide information about the accuracy of the gradients. Based on these gradient uncertainties, the optimization step size can be optimally selected to avoid overshoots or similar phenomena. Furthermore, the validity and quality of the 16p model can be continuously checked using the uncertainties of the 16p model, the sensor data, and the communication. This can be used to provide a cost function for retraining the 16p model or to signal necessary maintenance / repair.

[0046] This results in the following advantages: The interface component in distributed components of a machine learning system can now also be applied to processes with physical material flows, so that industrial process chains that operate on physical material flows can also be optimized. These optimization methods continue to protect trade secrets, which is made possible by the decentralized optimization method using the interface component in combination with decentralized condition assessment. According to exemplary embodiments, the machine can be assessed for wear or degeneration (e.g., abrasion, clogging, etc.), namely because the existing model 16p of the digital twin no longer perfectly matches the real process.Since derivations regarding the models are recorded, a model that no longer matches the actual behavior of the process step can be marked as a model or component that needs to be adapted. The consequence of a model that no longer fits is either the signaling of maintenance (generally deriving maintenance information) or a retraining of the model. For example, if a component 20b is replaced by a very similar component 20b', the model can be updated using this approach.

[0047] According to another embodiment, it is also conceivable to exchange uncertainties in the gradients of the digital twins as part of the external data 12d in order to prevent overshoot and instabilities in the optimization. Depending on the uncertainty, the step size can be adjusted during the optimization.

[0048] According to embodiments, an API consisting of an interface component can be used for distributed components of a machine learning system if the API is extended by communication options for the states of the digital twins. According to embodiments, the digital twin can be configured to provide gradient information. The gradient information is sufficient to enable decentralized data fusion for state estimation of the digital twins. In this respect, it is sufficient to exchange this gradient information as external data. According to embodiments, the processor is configured to create differentiability of the decentralized data fusion in order to enable backpropagation of the gradients. Differentiability here means that the influence of external information, e.g.external data in the form of backpropagated gradients or external cost functions, to directly calculate the state of the digital twin. As already explained above, uncertainties in the gradients are captured by the consistent statistical analysis and can be taken into account during optimization in the form of adaptation of the optimization step size by the processor 16. The combination of interface component 10 for distributed components 20 of a machine learning system with a differentiable digital twin and the integration of state communication into the API represents a particular advantage. The consequence is a differentiable, decentralized data fusion for state estimation of the digital twins. The optimization via the interface components 10 for distributed components 20 of a machine learning system can use conventional optimization algorithms and is executed locally.Decentralized data fusion is used for this purpose.

[0049] As already explained above, the main application is for processes that have not yet been or only partially digitized, such as those found in production plants or material sorting plants. These plants can be expanded by the interface component 10 or several interface components 10 per component in a plant. According to embodiments, individual interface components can not only communicate within the internal network of the production plant / material sorting plant, but can also exchange information with external sources, e.g. with upstream or downstream processes. According to preferred embodiments, the interface components communicate in particular with the upstream and downstream interface components in the direction of material flow. According to embodiments, it would also be conceivable that external data (cf.12d) In addition to the gradients, cost functions are also provided to the individual interface components as a type of objective. The differentiable process model advantageously makes it possible to optimize the cost function, e.g., by adjusting process parameters. The background is that the processor can represent each process parameter in a differentiable manner, together with its influence on another parameter, such as a cost function. Adjusting these process parameters also causes other process parameters, e.g., other components. The individual interface component that performs an adjustment locally with regard to the cost function can then make the respective process parameter or a gradient of the respective process parameter (derivative of the process parameter) available to the other interface components as external data (backpropagation).

[0050] According to further embodiments, the processor of an interface component can be connected to a memory, in particular an internal memory, i.e., a memory integrated in the interface component, on which the decentralized process model is stored. The processor calls this process model and modifies it or uses it using the external data and / or component parameters. According to embodiments, this decentralized process model enables the modeling of the state of the respective distributed component, taking into account the internal parameters and also the adjacent external parameters, e.g., of upstream or downstream units. The multiple decentralized process models together represent a higher-level process, which can be referred to as an overall process or as part of an overall process.The decentralized process model represents a local, decentralized, and decentrally stored and used process model that, when viewed together with other decentralized process models, forms a higher-level process model. However, the corresponding model data is still stored decentrally. The combination of at least two interface components forms a system with a higher-level process model comprising at least two decentralized process models belonging to the respective distributed components.

[0051] According to exemplary embodiments, not only the models are stored locally, but also the status data or any sensor measurements are stored locally. This advantageously allows only information concerning physically exchanged material or energy flows to be exchanged with neighboring nodes. Thus, status detection can also be implemented not via a centralized cloud but in a distributed system.

[0052] Although some aspects have been described in the context of a device, it should be understood that these aspects also represent a description of the corresponding method, so that a block or component of a device can also be understood as a corresponding method step or as a feature of a method step. Analogously, aspects described in the context of or as a method step also represent a description of a corresponding block, detail, or feature of a corresponding device. Some or all of the method steps may be performed by (or using) a hardware apparatus, such as a microprocessor, a programmable computer, or an electronic circuit. In some embodiments, some or more of the key method steps may be performed by such an apparatus.

[0053] Depending on specific implementation requirements, embodiments of the invention may be implemented in hardware or software. The implementation may be performed using a digital storage medium, such as a floppy disk, a DVD, a Blu-ray Disc, a CD, a ROM, a PROM, an EPROM, an EEPROM, or a FLASH memory, a hard disk, or other magnetic or optical storage device storing electronically readable control signals that can interact or cooperate with a programmable computer system to perform the respective method. Therefore, the digital storage medium may be computer-readable.

[0054] Some embodiments according to the invention thus comprise a data carrier having electronically readable control signals capable of interacting with a programmable computer system such that one of the methods described herein is carried out.

[0055] In general, embodiments of the present invention can be implemented as a computer program product with a program code, wherein the program code is effective to perform one of the methods when the computer program product is run on a computer. The program code can also be stored, for example, on a machine-readable medium.

[0056] Other embodiments include the computer program for performing one of the methods described herein, wherein the computer program is stored on a machine-readable medium. In other words, one embodiment of the method according to the invention is thus a computer program that has program code for performing one of the methods described herein when the computer program is executed on a computer.

[0057] A further embodiment of the method according to the invention is thus a data carrier (or a digital storage medium or a computer-readable medium) on which the computer program for carrying out one of the methods described herein is recorded.

[0058] A further embodiment of the method according to the invention is thus a data stream or a sequence of signals that represents the computer program for carrying out one of the methods described herein. The data stream or the sequence of signals can be configured, for example, to be transferred via a data communication connection, for example, via the Internet.

[0059] A further embodiment comprises a processing device, for example a computer or a programmable logic device, which is configured or adapted to carry out one of the methods described herein.

[0060] A further embodiment comprises a computer on which the computer program for performing one of the methods described herein is installed.

[0061] A further embodiment according to the invention comprises a device or system designed to transmit a computer program for performing at least one of the methods described herein to a recipient. The transmission can be electronic or optical, for example. The recipient can be, for example, a computer, a mobile device, a storage device, or a similar device. The device or system can, for example, comprise a file server for transmitting the computer program to the recipient.

[0062] In some embodiments, a programmable logic device (e.g., a field-programmable gate array, an FPGA) may be used to perform some or all of the functionalities of the methods described herein. In some embodiments, a field-programmable gate array may cooperate with a microprocessor to perform any of the methods described herein. In general, in some embodiments, the methods are performed by any hardware device. This may be general-purpose hardware such as a computer processor (CPU) or method-specific hardware such as an ASIC.

[0063] The above-described embodiments are merely illustrative of the principles of the present invention. It is understood that modifications and variations of the arrangements and details described herein will be apparent to others skilled in the art. Therefore, it is intended that the invention be limited only by the scope of the following claims and not by the specific details presented in the description and explanation of the embodiments herein.

Claims

Patent claims 1. Interface component (10) of a distributed component (20a, 20b, 20c) of a network, wherein the network has a plurality of distributed components (20a, 20b, 20c), each with an interface component (10), wherein the interface component (10) comprises the following features: a processor (16) which is designed to determine a state (16z) of a decentralized process model (16p) associated with the respective distributed component (20a, 20b, 20c); a communication interface (12) which is designed to exchange external data (12d) for the process model (16p) associated with a further distributed component (20a, 20b, 20c) of the plurality of distributed components (20a, 20b, 20c); an internal interface (14) configured to receive component parameters for the process model (16p) associated with the distributed component (20a, 20b, 20c); wherein the state (16z) is determined based on the component parameters and / or the incoming external data (12d).

2. Interface component (10) according to claim 1, wherein each decentralized process model (16p) has a state (16z) which is differentiable with respect to all component parameters and / or all incoming external data (12d).

3. Interface component (10) according to one of the preceding claims, wherein the state (16z) together with the decentralized process model (16p) forms a digital twin of the respective component.

4. Interface component (10) according to one of the preceding claims, wherein the external data (12d) comprise a gradient of component parameters and / or a derivative of component parameters and / or a cost function.

5. Interface component (10), wherein the communication interface (12) is designed to propagate component parameters and / or process parameters back as outgoing external data (12d).

6. Interface component (10) according to one of the preceding claims, wherein the state (16z) is derivable with respect to the decentralized model.

7. Interface component (10) according to one of the preceding claims, wherein the decentralized model is derivable with respect to incoming external data (12d); and / or wherein the derivation of the incoming external data (12d) symbolizes an influence on the process model (16p).

8. Interface component (10) according to one of the preceding claims, wherein the processor (16) is designed to provide a derivation of the state (16z) according to the component parameters and / or the incoming external data (12d) and / or the process model (16p).

9. Interface component (10) according to one of the preceding claims, wherein the processor (16) is designed as part of the network for decentralized data fusion in order to effect a state estimation of a digital twin.

10. Interface component (10) according to one of the preceding claims, wherein the processor (16) uses inferential statistics, in particular Bayesian methods, to determine the state (16z).

11. Interface component (10) according to one of the preceding claims, wherein the processor (16) is designed to determine an optimization step size for optimizing the decentralized process model based on gradients or information about an uncertainty.

12. Interface component (10) according to one of the preceding claims, wherein the communication interface (12) is designed to receive a higher-level parameter and / or a cost function associated with the higher-level parameter.

13. Interface component (10) according to one of the preceding claims, wherein the component parameters comprise sensor parameters.

14. Interface component (10) according to one of the preceding claims with reference to claim 6, wherein the processor (16) is designed to detect a deviation of the decentralized model.

15. The interface component (10) according to claim 14, wherein the processor (16) is configured to update and / or redefine the decentralized process model (16p) when a deviation of the decentralized model is detected; and / or to determine information regarding a maintenance requirement of the distributed component (20a, 20b, 20c) based on the deviation.

16. Interface component (10) according to one of the preceding claims, wherein the processor (16) is designed to perform an optimization of the decentralized model based on the component parameters and / or the incoming external data (12d).

17. Interface component (10) according to one of the preceding claims, wherein a plurality of interface components (10) of the network are interconnected according to a material flow (22) between the plurality of distributed components (20a, 20b, 20c) and the distributed component (20a, 20b, 20c); and / or wherein the processor (16) is configured to determine the state (16z) based on incoming external data (12d) from further distributed components (20a, 20b, 20c) taking into account a material flow (22) between the plurality of distributed components (20a, 20b, 20c) and the distributed components (20a, 20b, 20c).

18. The interface component according to one of the preceding claims, wherein the interface component has an internal memory on which the decentralized process model is stored.

19. A network comprising a plurality of distributed components (20a, 20b, 20c), each having an interface component (10), wherein the interface component (10) comprises the following features: a processor (16) configured to determine a state (16z) of a decentralized process model (16p) associated with the respective distributed component (20a, 20b, 20c); a communication interface (12) configured to exchange external data (12d) for the process model (16p) associated with a further distributed component (20a, 20b, 20c) of the plurality of distributed components (20a, 20b, 20c); an internal interface (14) configured to receive component parameters for the process model (16p) associated with the distributed component (20a, 20b, 20c); wherein the state (16z) is determined based on the component parameters and / or the incoming external data (12d).

20. The network of claim 19, wherein at least two interface components of the plurality of distributed components interact to form a higher-level process model comprising the respective decentralized process models.

21. A production plant or material sorting plant comprising a network according to claim 19 or 20.

22. Method for monitoring the status of at least one distributed component (20a, 20b, 20c) with an interface component (10) in a network comprising several distributed components (20a, 20b, 20c), comprising the following steps: Determining a state (16z) of a decentralized process model (16p) associated with the respective distributed component (20a, 20b, 20c); Exchanging external data (12d) for the process model (16p) associated with another distributed component (20a, 20b, 20c); Obtaining a component parameter for the process model (16p) associated with the distributed component (20a, 20b, 20c); wherein the state (16z) is determined based on the component parameters and / or the incoming external data (12d).

23. The method according to claim 21, wherein the method comprises the step of mapping a digital twin associated with the distributed component (20a, 20b, 20c).

24. Computer program for carrying out the method according to claims 22 and 23.

Images