Modularized model predictive control for industrial plants
A method for generating a dynamic model of an industrial plant having: a plurality of physical processes that are dependent such that an outcome of at least one first process is fed into at least one second process; a plurality of low-level controllers, each controller acting upon at least one physical process such that at least one process variable of the at least one physical process is controlled to match a set-point of the low-level controller; and a plurality of sensors, each sensor measuring at least one process variable of one of the physical processes, and/or of the plant as a whole, the set-points of the low-level controllers and current values of the process variables measured by the sensors being the inputs of the model, and predicted future values of the process variables that are likely to result from applying the set-points to the low-level controllers being the outputs.
1. A method for generating a dynamic model of an industrial plant having: a plurality of physical processes that are dependent such that an outcome of at least one first process is fed into at least one second process; a plurality of low-level controllers, each controller acting upon at least one physical process such that at least one process variable of the at least one physical process is controlled to match a set-point of the low-level controller; and a plurality of sensors, each sensor being configured to measure at least one process variable of one of the physical processes, and/or of the plant as a whole, the set-points of the low-level controllers and current values of the process variables measured by the sensors being the inputs of the model, and predicted future values of the process variables that are likely to result from applying the set-points to the low-level controllers being the outputs of the model, the method comprising:
obtaining a division of a representation of the plant into a plurality of sub-units corresponding to one or more physical process modules of a modular enabled automation solution at the industrial plant, an input-output behavior of each sub-unit being characterized by respective subsets of the set-points, subsets of the current values of process variables, and subsets of the future predicted values of process variables;
obtaining a dynamic sub-model for each sub-unit based on a digital representation of the respective one or more physical process modules of the obtained sub-unit, inputs of the sub-model comprising the subset of set-points and the subset of current values of process variables, and outputs of the sub-model comprising the subset of future predicted values of process variables;
obtaining, from a layout of the plant, and/or from an existing composition of the plant from physical process modules, dependencies between the sub-units; and
combining the sub-models into the model of the plant according to the dependencies for the manufacture of a given product, wherein the physical process modules of the model are linked together based on a temporary linking for the time during which a configuration of the plant for a process that manufactures the given product shall be used and until the sub-models are recombined for a configuration of the plant that is adapted to a further process that manufactures a further product.
2. The method of claim 1 , wherein a dynamic model provided by the module is chosen as the sub-model for the at least one sub-unit.
3. The method of claim 1 , further comprising:
obtaining information regarding the inputs and outputs of the sub-unit, and/or regarding a dynamic model of the sub-unit, and/or further information about the operation, interfaces, simulation, automation, and parameters of the sub-unit from a Module Type Package (MTP) corresponding to the sub-unit.
4. The method of claim 1 , wherein the obtaining the division into sub-units comprises searching for sub-units of which there are multiple occurrences in the plant, and
wherein each such sub-unit comprises at least two low-level controllers and at least two sensors.
5. The method of claim 4 , wherein the searching is focused on sub-units of a specific size, and if no, or no more, sub-units of the specific size are found, the searching is continued with a reduced size.
6. The method of claim 1 , wherein, in response to determining that there is a lesser requirement for an accuracy of the sub-model of a second sub-unit than there is for an accuracy of the sub-model of a first sub-unit, the second sub-unit is chosen to be larger than the first sub-unit.
7. The method of claim 1 , wherein at least one boundary of a sub-unit is chosen to coincide with a boundary between different areas of responsibility in the plant.
8. The method of claim 1 , wherein the dynamic model of the industrial plant comprises a first dynamic model for producing the given product, the method further comprising:
combining a subset of sub-models of the plurality of sub-models, according to the dependencies, into a second dynamic model of the industrial plant for producing the further product,
wherein the second dynamic model is an alternative to the first dynamic model.
9. A non-transitory computer-readable medium having processor-executable instructions stored thereon, wherein the processor-executable instructions, when executed by one or more processors, facilitate performance of the method of claim 1 .
10. A method for generating a dynamic model of an industrial plant having: a plurality of physical processes that are dependent such that an outcome of at least one first process is fed into at least one second process; a plurality of low-level controllers, each controller acting upon at least one physical process such that at least one process variable of the at least one physical process is controlled to match a set-point of the low-level controller; and a plurality of sensors, each sensor being configured to measure at least one process variable of one of the physical processes, and/or of the plant as a whole, the set-points of the low-level controllers and current values of the process variables measured by the sensors being the inputs of the model, and predicted future values of the process variables that are likely to result from applying the set-points to the low-level controllers being the outputs of the model, the method comprising:
obtaining a division of a representation of the plant into a plurality of sub-units corresponding to one or more physical process modules of a modular enabled automation solution at the industrial plant, an input-output behavior of each sub-unit being characterized by respective subsets of the set-points, subsets of the current values of process variables, and subsets of the future predicted values of process variables, wherein the division is obtained from a Module Type Package (MTP) that is a digital twin of the physical process module that contains a dynamic model of the respective sub-unit that corresponds to the module, information regarding the inputs and outputs of this sub-unit, and/or information about the operation, interfaces, simulation, automation and/or parameters of the respective sub-unit;
obtaining a dynamic sub-model for each sub-unit based on a digital representation of the respective one or more physical process modules of the obtained sub-unit, inputs of the sub-model comprising the subset of set-points and the subset of current values of process variables, and outputs of the sub-model comprising the subset of future predicted values of process variables;
obtaining, from a layout of the plant, and/or from an existing composition of the plant from physical process modules, dependencies between the sub-units; and
combining the sub-models into the model of the physical process modules of the plant according to the dependencies for the manufacture of a given product.
11. The method of claim 10 , wherein the obtaining the division into sub-units comprises searching for sub-units of which there are multiple occurrences in the plant, and
wherein each such sub-unit comprises at least two low-level controllers and at least two sensors.
12. The method of claim 11 , wherein the searching is focused on sub-units of a specific size, and if no, or no more, sub-units of the specific size are found, the searching is continued with a reduced size.
13. The method of claim 10 , wherein, in response to determining that there is a lesser requirement for an accuracy of the sub-model of a second sub-unit than there is for an accuracy of the sub-model of a first sub-unit, the second sub-unit is chosen to be larger than the first sub-unit.
14. The method of claim 10 , wherein at least one boundary of a sub-unit is chosen to coincide with a boundary between different areas of responsibility in the plant.
15. The method of claim 10 , wherein the dynamic model of the industrial plant comprises a first dynamic model for producing the given product, the method further comprising:
combining a subset of sub-models of the plurality of sub-models, according to the dependencies, into a second dynamic model of the industrial plant for producing the further product,
wherein the second dynamic model is an alternative to the first dynamic model.