Method for controlling a continuous granulation and drying process, and equipment and system therefor

Abstract

The present invention relates to a method for controlling a continuous granulation and drying process, as well as an equipment and a system therefor, wherein a model takes into account the combination of granulation using a granulator and subsequent drying using a dryer, and based on state parameters of the equipment, the model is used to determine control parameters or predicted formulation parameters, and / or the model comprises a static part, whereby a base value for each control parameter or predicted actual formulation parameter is determined or has been determined using the state parameters, and the model comprises a dynamic part, whereby the base values ​​are optimized by prediction.

Description

[Technical Field]

[0001] The invention relates to a method for controlling an installation, as well as to an installation according to the preamble of claim 14, and to a system comprising the installation. [Background technology]

[0002] The background of the present invention is mainly the production of pharmaceutical dosage forms, especially tablets, capsules or granules.However, in principle, the present invention can also be used in other technical fields, especially when a formulation is produced with specified / preset or pre-settable attributes related to the particles of the formulation, such as specified / preset particle size distribution and / or moisture.This is the case, for example, when the formulation is tableted.This is certainly particularly preferred in the pharmaceutical field, but in principle it can also be used in connection with detergents, food, etc.

[0003] The present invention particularly relates to a method and an installation for the preferably continuous production of a pharmaceutical preparation from raw materials, wherein the installation is particularly preferably controlled in the proposed manner and / or configured in such a way that a pharmaceutical preparation is produced from the raw materials having specified / preset or pre-determinable properties related to the particles of the pharmaceutical preparation, such as homogeneity and / or particle distribution, preferably specified / pre-set or pre-determinable (relative) moisture.

[0004] The combination of a granulator and a dryer has proven advantageous for the production of formulations from raw materials. Thus, an intermediate product (granule) with a specified / preset and / or pre-determinable particle distribution can first be achieved in a granulator starting from the raw materials, after which the intermediate product can be conditioned to a specified / preset or pre-determinable (relative) moisture by means of a dryer.

[0005] One difficulty here is the fact that various properties of the intermediate product affect drying and therefore formulation.

[0006] In principle, fluidized bed granulators are known. In the solutions disclosed there, a specific particle distribution with respect to relative moisture and lining is achieved in the same step. However, solutions based on fluidized bed granulators have drawbacks with respect to the reliable production of accurate particle size distributions and moisture, and are furthermore only suitable for certain raw materials.

[0007] In comparison, the use of a granulator followed by a separate dryer has proven to be more advantageous, since the components can be used separately here, and if appropriate, more precise particle size distributions and relative moisture contents can be achieved, preferably at least substantially independently of each other. However, the proposed method can also be advantageous in principle for controlling a fluidized bed granulator.

[0008] Granulator-dryer combinations that are initially batch-operated are also known in principle, in which a batch of raw materials undergoes a first production step, and then the entire batch undergoes another second production step before the result, i.e. the formulation, emerges from the production process.

[0009] In contrast, in the continuous manufacturing process that preferably forms the basis of the present invention, after the start-up stage, raw materials are simultaneously supplied while the previously supplied raw materials undergo the manufacturing process, and previously supplied raw materials that have already completely undergone the manufacturing process are also removed. Thus, in a continuous process, raw materials are simultaneously supplied and the resulting product is removed in the form of a formulation.

[0010] The present invention relates to continuous manufacturing of pharmaceutical preparations and / or continuous methods and equipment and / or systems having such equipment, preferably as opposed to batch processes. The advantages of continuous processes are: - Simple scalability (scaling over time and throughput) - Small space required for the equipment - Less equipment downtime compared to batch equipment - High degree of automation is possible - Higher product quality Summary of the Invention

[0011] Against this background, the present invention aims to provide methods, as well as equipment and systems, that can improve the process of producing pharmaceutical preparations from raw materials, in particular with regard to the reliable and consistent realization of attributes such as particle size distribution and moisture.

[0012] This object is achieved by a method according to claim 1, an installation according to claim 14 or a system according to claim 15. Advantageous developments are the subject matter of the dependent claims.

[0013] On the other hand, the present invention relates to a method for controlling equipment for producing a pharmaceutical preparation from raw materials, the production including processing the raw materials using a granulator and drying an intermediate product produced from the raw materials using the granulator in a dryer.

[0014] In a first variant of the invention, the control parameters of the equipment are determined on the basis of a model, where the model takes into account pre-set target formulation parameters that represent the desired properties of the formulation produced or to be produced.

[0015] The desired properties of the formulation produced or to be produced are in particular the target grain characteristics and moisture content.

[0016] Here, variably preconfigurable target formulation parameters can be passed to the model for determining the control parameters, however, the target formulation parameters can alternatively or additionally be or have been taken into account when creating the model.

[0017] A control parameter is or represents a manipulated variable for controlling an actuator of the plant.

[0018] The state parameters of the equipment are determined and processed by the model, and for this purpose they are preferably passed to the model and used by the model to determine the control parameters.

[0019] Each of the state parameters represents a state of the equipment that affects production and is preferably a sensor value.

[0020] Furthermore, raw material parameters are taken into account by the model. In particular, variably preconfigurable raw material parameters are passed to the model for the model to determine the control parameters. However, the raw material parameters can alternatively or additionally be or have been at least partially taken into account when forming the model.

[0021] The raw material parameters describe the attributes of the raw material, in particular the moisture and / or grain properties.

[0022] In one embodiment, the dryer is coupled to the granulator in such a way that the intermediate product is automatically conveyed from the granulator into the dryer without interruption. According to the proposal, the model considers the continuously performed combination of granulation using the granulator and subsequent drying using the dryer.

[0023] In a second aspect, which can be combined with the first aspect, the model has a static part (whereby a base value for each control parameter is determined) and a dynamic part (whereby the base value is optimized by prediction).

[0024] In a second variant of the invention, the actual formulation parameters are predicted based on a (same or different) model.

[0025] Here, the model takes into account preset or presettable control parameters in this manner. Preferably, variably presettable control parameters (by which the model predicts the actual formulation parameters) are passed to the model. However, the control parameters can alternatively or additionally be or have been at least partially taken into account when forming the model.

[0026] Actual production parameters represent the actual properties of the formulation produced or to be produced, in particular the actual grain properties and moisture.

[0027] As already in the first variant, the state parameters are determined and processed by the model, and the raw material parameters are taken into account by the model.

[0028] Furthermore, as already in the previous variant, in one aspect the dryer is coupled to the granulator in such a way that the intermediate product is automatically conveyed from the granulator into the dryer without interruption. According to the proposal, the model takes into account the continuously performed combination of granulation using a granulator and subsequent drying using a dryer.

[0029] In a second aspect, which can be combined with the first aspect, the model also has, in a second variant, a static part (whereby a base value for each control parameter is determined) and a dynamic part (whereby the base value is optimized by prediction).

[0030] In the proposed method, raw material parameters are therefore initially preset or taken into account in order to control or support the control of the equipment that produces the formulation from the raw material. The raw material parameters describe the state of the raw material, in particular its moisture content and / or particle properties such as particle size distribution.

[0031] Raw materials in the sense of the present invention are preferably granulatable substances, i.e. substances that can be processed by a granulation process to form granules. The raw materials are very particularly preferably powders or granules, the grain properties of which can be changed by the granulation process.

[0032] Furthermore, the raw material is preferably a substance mixture, i.e., an at least substantially homogeneous mixture of different, preferably solid, components. These components may comprise an active substance, in particular a pharmacologically or other active substance, a filler, and / or a disintegrant. In particular, the raw material is an at least substantially homogeneous powder mixture.

[0033] In the manufacture of drug products from raw materials, the material parameters of the raw materials passing through the equipment, through its intermediate products, and into the final product (drug product) are preferably not determined by off-site sampling and analysis, but instead measurable parameters are used exclusively in ongoing, continuous operations.

[0034] Either no material parameters are determined at all, or at most in-line measurable material parameters, such as measurements from non-contact measuring methods, reflection and / or transmission measurements, especially those using infrared radiation, e.g., as an indicator of material moisture. Measurement of particle size distribution in the production process is preferably omitted, at least for intermediate products.

[0035] For the proposed control, control parameters of the equipment can be determined, which are or represent manipulated variables for controlling actuators of the equipment. The equipment can then be controlled with these control parameters, preferably by controlling different actuators of the equipment with the control parameters to influence the raw materials and / or intermediate products.

[0036] Alternatively or additionally, to support control, actual formulation parameters are predicted that represent the properties of the formulation produced under specified / predefined boundary conditions, which can be based on predefined or predefinable control parameters.

[0037] In other words, actual formulation parameters can be predicted and preferably output, preferably by manual provision or input of control parameters, which can then be compared with the target formulation parameters by a user (again, preferably manually) and various control parameters identified / provided to match the predicted actual formulation parameters to the target formulation parameters. The various control parameters are then preferably used as the basis for controlling the equipment.

[0038] Preferably, at least one granulator driver and the supply of desiccant, in particular (conditioned) air, are controlled by the control parameters. In addition, the raw material supply devices, the liquid injection during granulation, one or more temperature control devices of the granulator, the desiccant conveying device for setting the desiccant flow rate, and / or the temperature control device for controlling the desiccant temperature can be controlled by the control parameters.

[0039] To control and / or support the control of the equipment, state parameters of the equipment, in particular one or more sensor values, are determined, preferably each sensor value representing an equipment state that affects the process of producing a pharmaceutical product from raw materials. These include, in particular, temperature, pressure, pressure difference, torque, and / or volumetric flow rate. However, other parameters and / or sensor values ​​describing the equipment state are also conceivable.

[0040] However, the state parameters preferably do not, or at least do not directly, describe the material properties of the raw materials or the intermediate products (granulations) or final products (formulations) formed therefrom.

[0041] In any case, it is preferable that the particle size distribution, size, shape, density, and / or active ingredient content of the raw material or intermediate products formed therefrom are not determined in the continuously performed granulation and drying processes. In this respect, the present invention adopts a completely different approach from the prior art. In contrast, the characteristics of the raw material can be determined in advance, and the characteristics of the final product, i.e., the formulation, can be determined for verification and / or model formation after completion.

[0042] Target formulation parameters, which represent desired properties of the formulation produced or to be produced, in particular one or more physical properties such as particle size, particle size distribution, particle shape, or density, and / or moisture content of the formulation, are preferably preset for control.

[0043] Furthermore, for model generation and / or validation, preferably, actual formulation parameters that describe the actual properties of the formulation that have been produced or are to be produced, in particular one or more physical properties such as particle size, particle size distribution, particle shape, or density, and / or the moisture content of the formulation, are determined by characterization of the formulation.

[0044] Finally, it is provided that the control parameters are determined based on the model. Then, in the process of manufacturing the formulation, the equipment can be controlled preferably using the control parameters determined by processing the state parameters with the model.

[0045] Preferably, the (measured) actual formulation parameters are used to derive and / or define the model, however, the (measured) actual formulation parameters are preferably not used as a basis for controlling equipment in an ongoing process.

[0046] Therefore, the control parameters are preferably not determined by processing and / or deriving them from (measured) actual formulation parameters. This is because it has surprisingly been found that deriving the control parameters from the (measured) actual formulation parameters starts too late. If the (measured) actual formulation parameters deviate from the target formulation parameters in an ongoing process, a significant number of rejects will already be pre-programmed. However, the object of the present invention is to avoid such rejects. For this purpose, it is rather preferred to control or configure the equipment (at least substantially) independently of the (measured) actual formulation parameters.

[0047] The (measured) actual formulation parameters are preferably used to determine the model and / or to derive a scheme in the form of a model, whereby the control parameters are derived from the state parameters of the equipment in the process.

[0048] The model preferably has a machine learning based structure, in particular a neural network.

[0049] As already mentioned, in a first aspect of the invention, the installation comprises a granulator for processing raw materials to form an intermediate product, and a dryer, the dryer being coupled to the granulator such that the intermediate product is automatically conveyed from the granulator into the dryer without interruption.

[0050] Here, the model considers, and in particular describes, the combination of granulation using a granulator and subsequent drying using a dryer. Using the model, the control parameters, in particular the base values ​​of the control parameters, can be determined based on the equipment state parameters, preferably raw material parameters.

[0051] Alternatively or additionally, predicted (and therefore unmeasured) actual formulation parameters are determined based on a model, which may be a different model from the model determining the control parameters. The predicted actual formulation parameters can be output to allow for (manual) comparison with target formulation parameters and, if appropriate, (manual or automatic) adjustment of the control parameters. In this case, the control parameters are preferably preset relative to the model.

[0052] Optionally, one or more in-line measurable properties of the intermediate product may additionally be taken into account, in particular one or more parameters that are optically measurable on the intermediate product, however preferably measurements that require sampling, analysis separate from the facility or interruption of the manufacturing process for their determination are avoided.

[0053] Therefore, in addition to the equipment state parameters, it is not excluded that certain (physical) in-line measurable properties (actual formulation parameters) of the raw materials and / or intermediate products and / or formulations (e.g., the moisture and / or particle size distribution of formulation 7 or variables determined therefrom) may be determined or used for closed-loop control and / or as input variables for the model, especially if in-line measurements for the determination are possible without interrupting the continuous granulation and drying.

[0054] Preferably, at most or only measurements of the particle size distribution of the formulation are provided here, but not measurements of the intermediate product.Thus, the equipment can have a sensor for determining attributes describing the particles of the formulation, such as particle size distribution, but preferably only downstream of the formulation outlet for dispensing the formulation after drying the intermediate product.The sensor provided in the equipment, or as part of a system with the equipment, is an in-line probe that performs local filter anemometry, in particular for particle size measurement.However, other principles are also possible here.

[0055] Preferably, the moisture content is determined from the formulation, but alternatively or additionally, it can also be determined from an intermediate product. One or more sensors can be provided for this purpose. In particular, these are optical sensors, particularly preferably infrared radiation-based sensors. In this context, sensors based on near-infrared radiation, particularly the NIR-2 spectrum at wavelengths between 860 and 1040 nm, have proven particularly advantageous.

[0056] It has proven particularly advantageous if, independently of the granulator-dryer combination and in conjunction with the model formation of the granulator-dryer combination, the model has a static part (whereby base values ​​of each control parameter and / or predicted actual formulation parameter are determined from the raw material parameters and state parameters) and if the model also has a dynamic part (whereby the base values ​​are optimized or can be optimized by prediction).

[0057] The static part of the model can be defined by determining model parameters that correspond to each other and are determined, in particular measured, during a plurality of respective stationary states (steady states, at least substantially static states) of the installation.

[0058] The static part of the model can be used to determine baseline values ​​of control parameters from state parameters and / or to predict actual formulation parameters. The static part of the model is preferably provided as an invariant.

[0059] The dynamic part of the model can be used to continuously adjust the model's behavior. For example, equipment behavior changes over time due to wear, material expansion, aging, etc. Instead of taking such or similar transient effects into account directly by changing the values ​​of the control parameters or predicted actual formulation parameters (particularly by scaling and / or correcting by adding / subtracting correction terms), the model is adjusted according to the proposal so that the model generates corrected control parameters and / or actual formulation parameters from the state parameters.

[0060] One or more (different, if appropriate) of the parameters (such as state parameters) are processed by the model, where it may be provided that these parameters form input values ​​for the model and thus in particular for the static part of the model and / or the dynamic part of the model.

[0061] The model, and in particular the static part of the model, may comprise an artificial neural network: parameters are passed to nodes in the input layer of the artificial neural network, which may use these to generate values ​​at nodes in the output layer of the artificial neural network.

[0062] The values ​​generated by the nodes of the neural network in the static part of the model at the output layer can be optimized with the dynamic part of the model, for example by processing them taking into account the same and / or other parameters.

[0063] A further aspect of the invention, which can also be realized independently, relates to a computer program product or a computer-readable storage medium comprising instructions that, when the program is executed by a computer, cause the computer to perform a method according to one of the described aspects.

[0064] A further aspect of the present invention, which can also be implemented independently, relates to an apparatus for producing a pharmaceutical preparation from raw materials, the production including processing the raw materials using a granulator and drying the intermediate product produced using the granulator in a dryer.

[0065] The equipment has sensors for detecting equipment state parameters, each representing an equipment state that affects production. The equipment has actuators for directly or indirectly acting on raw materials. Furthermore, the equipment has control devices for controlling the actuators using control parameters, and target formulation parameters are preset or pre-settable for the control devices and models, based on which the control parameters can be determined.

[0066] The granulator is coupled to the dryer so that the intermediate product is automatically conveyed from the granulator into the dryer without interruption, the model considers a combination of granulation using the granulator and subsequent continuous drying using the dryer, and the equipment is configured such that the control device determines control parameters using the model based on state parameters of the equipment.

[0067] Alternatively or additionally, the model includes a static portion, and the control device is configured to determine a base value for each control parameter from the raw material parameters and the state parameters using the static portion, and the model includes a dynamic portion, and the control device is configured to optimize the base values ​​using the dynamic portion by prediction. The equipment is then preferably controlled with the resulting optimized base values ​​and / or the equipment is controllable using the base values.

[0068] A further aspect of the invention, which can also be realized independently, relates to a system comprising the proposed installation, a device for forming a raw material from a plurality of components, preferably powders, preferably by sieving, and / or a device for further processing of the formulation, preferably tableting.

[0069] A formulation in the sense of the present invention is a substance that passes through the equipment and is modified by the equipment in terms of its physical properties.Therefore, the formulation is preferably the product of the combination of granulation and drying.Further processing of the formulation that is preferably present as granulation (preferably dried and / or adjusted in terms of its (relative) moisture content), for example by tableting, is not excluded.

[0070] The raw material in the sense of the present invention is preferably a substance that is fed into the equipment and / or granulator to change the physical properties of the raw material.The raw material is preferably an active ingredient-filler mixture.However, this is not essential.The raw material can already be pre-processed, for example, by at least substantially homogeneously mixing a powder with another powder or another substance (one of which can be or can contain an active substance).The active substance is preferably a pharmacologically active substance.

[0071] A granulator in the sense of the present invention is preferably a device for mechanically processing raw materials to change their physical properties. Particularly preferably, the granulator converts the raw materials into granules, in particular into coarse (granular) powder. For this purpose, the raw materials can be fed to the granulator as powder, in order to process them into coarser or finer granular powder. The intermediate product is preferably solid.

[0072] Preferably, the granulator conveys the raw materials while acting on them. The granulator preferably generates pressure and / or friction within the raw materials, accompanied by temperature control / heating and / or moisture supply. Particularly preferably, the granulator changes the particle size, particle size, or particle size distribution of the raw materials. In the present invention, the granulated product is an intermediate product, which is further processed as follows.

[0073] The granulator can have an extruder or can be formed by an extruder. In particular, it is a screw extruder, preferably a twin-screw extruder, or the granulator has or is similar to such a screw extruder. However, other solutions are also conceivable here.

[0074] The granulator is preferably a screw granulator, for example a twin-screw granulator. In a screw granulator, the raw material is conveyed and processed by a screw shaft rotating around the rotation axis. For this purpose, the screw shaft can have screw flights with different pitches and / or processing configurations. A twin-screw granulator is provided with two screw shafts arranged parallel and / or interlocking to provide conveying and processing. In principle, other concepts can also be used here. Conveying preferably results in extrusion. Thus, the screws form one or more extruders and / or achieve extrusion of the raw material.

[0075] The granulator is preferably a granulator for wet granulation. For this purpose, the granulator may have an injection of a liquid to moisten the raw materials, which are processed by the granulator to form an intermediate product. The granulator may have openings, in particular nozzles and / or valves, for adding a liquid to the raw materials to (temporarily) increase their moisture content, in particular water, ethanol, isopropanol, and / or mixtures thereof.

[0076] A dryer in the sense of the present invention is a device for reducing the (relative) moisture and / or water content of a substance, in this case a device for reducing the (relative) moisture and / or water content of raw materials that are processed in a granulator to become an intermediate product. The dryer is preferably a device that draws water from the substance / intermediate product. It is preferably a device that brings the substance / intermediate product into contact with a desiccant that draws water from the substance.

[0077] Preferably, the desiccant is air or other gas with a relative humidity that allows the absorption of water from the material. Preferably, the desiccant is temperature-controlled, particularly above ambient temperature. Thus, the desiccant is particularly preconditioned, warm and / or dry air, also called process air. However, in principle, the desiccant can also be other, particularly inert, gases or other, preferably gaseous, desiccants.

[0078] The dryer is preferably a fluidized bed dryer. A fluidized bed dryer in the sense of the present invention is a device that generates a cushion of air or gas (process air) for the material to be dried, in this case the intermediate product / granulate. In this context, the air or gas is preferably a gaseous desiccant.

[0079] The desiccant, preferably gaseous, is fed from the intermediate product / granules to the dryer bed, preferably through a perforated distributor plate. The desiccant, preferably gaseous, flows through the bed at a velocity such that the particles of the intermediate product / granules remain fluidized despite their weight. The fluidized particles of the intermediate product / granules form a fluidized bed, where they are preferably dried by the gaseous desiccant.

[0080] In the fluidized bed, bubbles can form and collapse, promoting intensive particle movement. In this state, the solids behave like a free-flowing boiling liquid. Extremely high heat and mass transfer rates result from the intimate contact between the individual particles and the preferably gaseous desiccant. However, while a fluidized bed dryer has proven particularly advantageous in the context of the present invention, other dryer concepts can also be used in principle.

[0081] Sensors for detecting state parameters of an installation, each representative of a state of the installation that influences production, in the sense of the present invention are preferably sensors that characterize the state of a device of the installation, i.e. devices that are oriented and configured to measure one or more state parameters.

[0082] A state parameter in the sense of the present invention may be or represent one or more of the following attributes: - the moment / torque and / or speed of the drivers, tools and / or conveying devices (in particular of propellers, turbines, screws / extruders), or their corresponding parameters, such as current consumption and / or rotational speed; Mass flow rates (of desiccant / process gas, raw material, agglomeration / granulation liquid, intermediate product, exhaust gas, final product / formulation) or corresponding parameters characterizing, for example, valve positions, flaps, rotor rotation speed, pressure differences, etc. Preferably the temperature of the parts of the equipment that come into direct or indirect contact with the raw material being processed, the flocculating / granulating liquid, or the desiccant, the temperature of the tools for processing the raw material or intermediate product, the temperature of the device for adding the flocculating liquid or introducing the desiccant

[0083] The state parameters preferably do not (directly) characterize any (physical or chemical) properties of the raw materials or the intermediate or final products formed therefrom and / or formulations. Thus, in particular, measured variables (characterizing chemical composition or particle size) of the raw materials or the intermediate or final products (formulations) formed therefrom are not considered state parameters of the equipment.

[0084] In principle, a distinction can be made between two groups of state parameters.

[0085] The first group includes state parameters that are at least substantially independent of the feedstock, in particular in that they are set directly by the actuators of the equipment. These are also referred to hereinafter as pre-settable and / or "feedstock-independent state parameters". Examples of this are pre-settable temperatures and rotation speeds. They can be used in particular for actuator control.

[0086] A second group of state parameters, which are influenced by interactions with the raw material or intermediate products formed therefrom, can be distinguished from the "raw material-independent state parameters". These will be referred to hereafter as "raw material-influenced state parameters". Examples of this are the torque generated as a function of the raw material consistency or the (relative) humidity of the exhaust air moistened during a drying process.

[0087] State parameters influenced by one or more raw materials are preferably used as input variables of the model. The state parameters processed by and / or passed to the model for this purpose are used by the model to determine control parameters and / or predict actual formulation parameters, and are therefore preferably state parameters influenced by raw materials.

[0088] For this purpose, the state parameters processed by and / or passed to the model and used by the model to determine the control parameters and predict the actual formulation parameters are preferably state parameters influenced by at least one raw material in the granulator, more preferably state parameters influenced by at least one raw material in each of the granulator and the dryer, in particular state parameters influenced by at least two raw materials in the granulator, and state parameters influenced by at least one, preferably at least two, raw materials in the dryer.

[0089] Preferably, one or more of the parameters that are not affected by the raw materials are used to perform closed-loop control of the corresponding actuator. Alternatively or additionally, one or more state parameters that are affected by the raw materials are used to determine control parameters, preferably by a model. The control parameters may be target specifications for closed-loop control of the actuator, and the actuator is closed-loop controlled based on them.

[0090] In-line measurable properties in the sense of the present invention are properties that can be measured concomitantly in a preferably uninterrupted, continuous manufacturing process.

[0091] Actuators for acting on raw materials in the sense of the present invention are preferably drivers of tools and / or conveying devices. These can be, for example, motors used to drive fans, turbines, conveyor belts, and / or screws, or temperature control devices, heaters, or coolers (e.g., cooling water conveying devices) for controlling the temperature of the housing parts of the granulator that carry the raw materials. Temperature control devices for controlling the temperature of process gases, in particular heaters or coolers such as heating registers, can also be actuators, since they act indirectly on the raw materials via the temperature of the desiccant / process gas.

[0092] Control devices for controlling the actuators in the sense of the present invention are preferably electronic components that influence the operation of the actuators, that control the motor rotation speed (of the screw driver and / or fan motor) and / or that open-loop or closed-loop control the temperature control devices and / or heaters of the granulator and / or desiccant.

[0093] Control parameters which are or represent manipulated variables for controlling an actuator in the sense of the present invention are preferably values ​​as specifications for the operation of the actuator, in particular one or more specified / preset motor rotation speeds (of the screw driver and / or fan motor) or the corresponding current consumptions, and / or specifications for open-loop or closed-loop control of temperature control devices and / or heaters, such as granulators and / or desiccants.

[0094] Feedstock parameters in the sense of the present invention are preferably parameters characterizing the physical properties of the feedstock, such as the particle size, particle size, particle size distribution, (relative) moisture and / or temperature of the feedstock.

[0095] The condition of a raw material in the sense of the present invention is preferably determined at least by its moisture content and / or particle size distribution, optionally supplemented by further properties or replaced by corresponding information that allows conclusions to be drawn or derived directly or indirectly about the moisture content and / or particle size distribution.

[0096] Target formulation parameters in the sense of the present invention are parameters that preferably describe the desired properties of the formulation produced or to be produced, in particular the particle size and / or particle size distribution and moisture, optionally supplemented by further properties or replaced by corresponding information that allows conclusions to be drawn or derived therefrom, directly or indirectly, about the moisture and / or particle size distribution.

[0097] Actual formulation parameters in the sense of the present invention are parameters measured on a formulation that are representative of the properties of the formulation produced or to be produced, in particular particle size, particle size distribution and moisture, optionally supplemented by further properties or replaced by corresponding information that allows conclusions to be drawn or derived therefrom, directly or indirectly, about the moisture and / or particle size distribution.

[0098] Actual formulation parameters can be predicted as an alternative to measurement, in which case they are referred to below as predicted actual formulation parameters.

[0099] A model in the sense of the present invention is preferably a representation (abstracted) preferably limited to essential characteristics. The model here is preferably a model of the equipment, preferably mathematically representing the characteristics of the equipment directly or indirectly through its influence on the raw materials and / or intermediate products. For this purpose, the model can have or be formed by a mathematical description of the granulator, dryer, and / or their influence on the raw materials and / or intermediate products. The model preferably makes it possible to determine control parameters and / or predict formulation parameters, starting from the equipment state parameters and, if appropriate, raw material parameters.

[0100] The coupling of the granulator to the dryer in the sense of the present invention preferably refers to a device for transferring the intermediate product into the dryer, and may include a conveying device for this transfer, such as a chute, conveyor belt, screw, etc. The coupling preferably has the effect that the intermediate product is continuously and automatically transported from the granulator outlet into the dryer without interruption and / or "in-line." For this purpose, the granulator may also have an outlet opening directly into the dryer. Here, the coupling is thus achieved by the granulator and / or by the conveying action of the granulator.

[0101] Baseline values ​​of control parameters in the sense of the present invention are control parameters which preferably serve as base settings and which are preferably determined or specified / pre-set independently of the measured properties of the final product (formulation). The base values ​​of the predicted actual formulation parameters are the starting point for the prediction.

[0102] The static part of the model in the sense of the present invention is preferably the part of the model that is based on empirical values ​​and with which the basic values ​​can be determined and / or predicted.

[0103] The dynamic part of the model in the sense of the present invention is preferably a part of the model that can be changed as a function of the state parameters of the equipment, i.e., during the ongoing operation of the equipment during the processing of raw materials into intermediate and final products of the formulation, in order to adjust the model accordingly to any (future) changes in the state of the equipment or process, so that the model preferably represents the behavior of the equipment and / or process for producing a pharmaceutical product from raw materials with sufficient accuracy. The dynamic part of the model makes it possible to optimize the base values, preferably by prediction. For this purpose, past developments can be taken into account.

[0104] Further aspects, advantages, and features of the present invention will emerge from the claims and the following description of preferred exemplary embodiments, with reference to the drawings. [Brief explanation of the drawings]

[0105] [Figure 1] FIG. 1 is a schematic cross-sectional view of the proposed installation. [Figure 2] FIG. 2 is a simplified block diagram-like illustration of components associated with the control of the installation of FIG. 1. [Figure 3] FIG. 1 is a simplified schematic diagram of an artificial neural network. [Figure 4] FIG. 1 is a schematic diagram of a past time course and a predicted time course. [Figure 5] FIG. 10 is a schematic illustration of the results of control with constant control parameters over time. [Figure 6] FIG. 1 is a schematic diagram of temporally offset predictions compared to each other. [Figure 7] FIG. 1 is a schematic diagram of process characteristics over time. [Figure 8] FIG. [Figure 9] It is a facility embedded in the system. DETAILED DESCRIPTION OF THE INVENTION

[0106] In the drawings, the same reference numerals are used for identical or similar parts, and the same or similar features and advantages may be achieved even if the description is not repeated for the sake of clarity.

[0107] 1 shows a schematic diagram of a proposed facility 1 for producing a formulation 2 from raw materials 3. The facility 1 is preferably configured to carry out a method for producing a formulation 2 from raw materials 3.

[0108] For the feedstock 3, feedstock parameters 4 are specified / predefined and describe the state of the feedstock 3, in particular the moisture, composition and / or grain characteristics such as particle size distribution.

[0109] The plant 1 is preferably configured to determine control parameters 5 of the plant 1. The control parameters 5 are or represent manipulated variables for controlling actuators 6 of the plant 1.

[0110] Furthermore, state parameters 7 of the equipment 1 are preferably determined or can be determined using the equipment 1, for example via sensors 8, in particular via sensor values ​​or variables derived therefrom as state parameters 7, each of which represents a state of the equipment 1 that affects production.

[0111] A control device 9 of the facility 1 may be used to control the actuators 6 of the facility 1 to control or influence the manufacturing process for forming the formulation 3 from the raw materials 2 .

[0112] Target formulation parameters 10 can be specified / preset, which represent desired properties of the formulation 2 produced or to be produced, particularly physical properties such as grain properties, which are particle size content and / or moisture.

[0113] The target formulation parameters 10 may be available stored and / or maintained in a database 10A that may be read by the control device 9 and / or the target formulation parameters 10 may be retrieved from the database 10A by the control device 9 and used for control.

[0114] Actual formulation parameters 11 can be provided or measured that represent the actual properties of the formulation 3 that has been or is to be produced, in particular physical properties such as particle size distribution and / or grain properties such as moisture. The actual formulation parameters 11 are preferably information about or have information about attributes of the formulation 2, the attributes being measured in a separate analytical process, in particular separate from the facility 1, preferably not in-line and / or not in real time / delayed.

[0115] The installation 1 , in particular the control device 9 , is preferably configured to determine the control parameters 5 based on a model 12 .

[0116] Alternatively or additionally, the proposed model 12 can be used to predict and / or output predicted actual formulation parameters 11 (expected under given boundary conditions) of the formulation 2, preferably independently of the control device 9 and / or without direct influence or preferably automatic control of the actuator 6. For this purpose, the model 12 can be used to predict one or more properties of the formulation 2 from one or more state parameters 7 and control parameters 5 and preferably raw material parameters 4.

[0117] In one embodiment of the present invention, the facility 1 comprises a granulator 13 for processing the raw material 2 to form an intermediate product 14 and a dryer 15 for the intermediate product 14 .

[0118] The dryer 15 is coupled to the granulator 13 so that the intermediate product 14 is conveyed automatically and / or without interruption from the granulator 13 to the dryer 15. Granulation and drying preferably form a generally continuous process.

[0119] A simplified block diagram-like illustration of the components involved in controlling the installation 1 is shown in FIG.

[0120] The control device 9 is preferably configured to control the installation 1 and / or the combination of the granulator 13 and the dryer 15. The control is preferably based on a model 12 describing the behavior of the installation 1 and / or the combination of the granulator 13 and the dryer 15 and making it possible to determine control parameters 5 for controlling actuators 6 of the installation 1 and / or the combination of the granulator 13 and the dryer 15 based on at least one or more state parameters 7. It is particularly preferred here that the model 12 takes into account the combination of granulation using the granulator 13 and subsequent drying using the dryer 15.

[0121] Using the model 12 , the control parameters 5 , in particular the initial base values ​​of the control parameters 5 , are preferably determined based on the state parameters 7 of the equipment 1 , and more preferably based on the raw material parameters 4 .

[0122] For this purpose, the model 12 may have a static part 12A in which a base value for each control parameter 5 is determined taking into account the raw material parameters 4 and based on the state parameters 7, and a dynamic part 12B in which the (respective) base values ​​can be adjusted, preferably optimized by prediction.

[0123] In other words, the static portion 12A of the model 12, preferably formed by machine learning, is used to determine base values ​​(base settings) of the control parameters 5, which can then be adjusted and finalized by the dynamic portion 12B of the model 12 for ultimate use in controlling the actuators 6.

[0124] The dynamic part 12B of the model 12 preferably realizes fine-tuning of the base values ​​and / or base settings determined in the static part 12A of the model 12. Here, the base values ​​of the control parameters 5 determined in the static part 12A are preferably independent of the transient behavior, i.e., the runtime behavior, for example, under the influence of environmental influences, tolerance changes, wear of the manufacturing process of the equipment 1 and / or the formulation 2.

[0125] In contrast, the adjustment values ​​determined using the dynamic portion 12B for the control parameters 5 and / or the settings and / or control parameters 5 adjusted accordingly take into account transient and / or run-time effects, for example via one or more forecasts, preferably taking into account past developments, in particular by time series forecasting.

[0126] For this purpose, adjustment values ​​of the base values ​​of the control parameters 5 and / or correspondingly adjusted settings and / or control parameters 5 can be determined, which preferably take into account a comparison between the predicted actual formulation parameters 11 and the target formulation parameters 10, and by dynamic adjustment of the base values ​​of the control parameters 5, in particular based on the current state parameters 7 at the time of execution, to bring the actual formulation parameters 11 closer to the target formulation parameters 10.

[0127] An advantageous feature of this procedure is that the reference value and the adjustment can be determined in different ways. Alternatively or additionally, at least the base value, preferably different from the adjustment, is determined by AI and / or machine learning methods. Furthermore, it is particularly preferred that the determination of the base value also depends on the state parameter 7. That is, the base value and the adjustment (correction term) for determining the adjusted and / or corrected base value are variable.

[0128] Thus, the base values ​​depend on the state parameters 7, but can be, and preferably are, independent of the past and predicted evolution of the state parameters 7. In contrast, the base values ​​of the control parameters 5 determined in the static part 12A of the model 12 can be variable depending on the (current) state parameters 7 and correction terms and / or adjustments determined in the dynamic part 12B of the model 12, which adjust and / or correct the base values.

[0129] In this case, the static part 12A of the model 12 preferably differs from the dynamic part 12B of the model 12 in that in the static part 12A, a base value is determined or determinable without predicting the influence of the state parameter 7 changing with changes in the control parameter 5 and / or past changes in the state parameter 7, while the dynamic part 12B takes into account the influence of the state parameter 7 changing with predictions obtained from changes in the control parameter 5 in the future and / or past changes in the control parameter 5 or the state parameter 7, more preferably taking into account previous changes.

[0130] In summary, the static part 12A of the model 12 differs from the dynamic part 12B of the model 12 in that it implements a different method for determining, in the case of the static part 12A, the base value of the control parameter 5 and the correction term for the base value of the control parameter 5, or the corrected / adjusted base value as the final control parameter 5, preferably using or from the state parameter 7 and preferably taking into account the raw material parameter 4.

[0131] The state parameters 7 taken into account as input variables by the model 12 are preferably state parameters 7 influenced by the raw materials. Here, the model 12 preferably uses at least one state parameter 7 influenced by the raw materials for each of the granulator 13 and the dryer 15.

[0132] Optionally, model 12 additionally considers as input variables (exclusively) in-line measurable formulation characteristic parameters 7K that describe physical or chemical properties of formulation 2 that are measurable in an uninterrupted, continuously run process.

[0133] Alternatively or additionally, the model 12 additionally considers as input variables (exclusively) in-line measurable intermediate product property parameters 7L describing physical or chemical properties of the intermediate product 14 that are measurable in an uninterrupted, continuously running process. The measurements can be performed using intermediate product property sensors 8L, in particular (NIR) sensors for determining an indication of the moisture content (water content) of the intermediate product 14.

[0134] In contrast, further consideration of other parameters is not excluded. Furthermore, although the definition of Model 12 may consider parameters that cannot be measured in-line, Model 12 preferably does not require parameters that cannot be measured in-line as input variables in ongoing operations.

[0135] The static part 12A of the model 12 may implement an AI and / or machine learning method, and the dynamic part 12B of the model 12 preferably implements another non-AI and / or non-machine learning based method and / or does not implement a neural network 32. However, it is not excluded that both different methods are AI and / or machine learning based, or both are not.

[0136] Surprisingly, it has been found that the closed-loop control of the installation 1 is particularly reliable and accurate, so that the actual formulation parameters 11 are particularly close to the target formulation parameters 10 when the aforementioned aspects are combined with one another. In summary, this is surprisingly highly advantageous in the following respects: Controlling or enabling control of the combination of granulator 13 and dryer 15, and / or Directly combining equipment components without external and / or non-in-line analysis of the intermediate product 14; and / or Excluding the use of analytical results for intermediate products 14 that are not measured or measurable in-line and therefore are not in a continuously running process, and / or for continuously moving intermediate products 14; and / or the basic values ​​of the control parameters 5 are / can be determined in the static part 12A of the model 12 using a first, preferably machine learning based method; and / or The base value of the control parameter 5 is determined / determinable without taking into account the predicted effect of the state parameter 7 that changes with the change of the control parameter 5 in the determination, and / or the dynamic part of the control parameters 5 is determined or determinable in the dynamic part 12B of the model 12, or when the base values ​​are corrected in the dynamic part or when the installation 1 is configured for this purpose, and / or the dynamic part of the control parameters 5 is determined or determinable in the dynamic part 12B of the model 12, preferably by a time series forecasting and / or non-machine learning based method separate from the method used or supported in the static part 12A of the model 12; and / or The dynamic part 12B of the model predicts the effect of changing the state parameters 7 as the control parameters 5 change, and takes this into account, or the equipment 1 is configured to do so; and / or the basic values ​​of the control parameters 5 are corrected and / or adjusted in the dynamic part, or correction terms are determined / determinable for this purpose, or the installation 1 is configured for this purpose; and / or ●The final control parameters 5 are the basic values ​​adjusted in the dynamic part, and equipment 1 is controlled by these control parameters 5.

[0137] As shown in FIG. 1, the facility 1 may have a number of different actuators 6 to act directly or indirectly on the raw materials 3 to form an intermediate product 14 and ultimately the formulation 2.

[0138] The control parameters 5 correspond to the respective actuators 6 and may be or correspond to manipulated variables of the actuators 6 for controlling the behavior of the actuators 6 .

[0139] The actuators 6 of the granulator 13 that are controllable via corresponding control parameters 5 include one or more of the following actuators 6: a feed conveyor driver 6A, preferably controllable by the feed conveyor driver control parameters 5A, for metering and / or conveying the raw material 3 to the granulator 13; and / or a granulator driver 6B, preferably controllable by granulator driver control parameters 5B, for driving the granulator unit 17 and / or one or more screws; and / or an injection device 6C, preferably controllable by injection device control parameters 5C, for injecting the granulation liquid 13B; and / or A granulator temperature control device 6D, controllable by granulator temperature control device control parameters 5D, preferably for controlling (heating and / or cooling) the temperature of the parts of the granulator 13 that come into contact with the raw material 3.

[0140] The actuators 6 of the dryer 15 that are controllable via corresponding control parameters 5 include one or more of the following actuators 6: a supply air conveyor driver 6E, preferably a fan, for supplying air 27 to the dryer 15, controllable by the supply air conveyor driver control parameters 5E; and / or a supply air temperature control device 6F, controllable by a supply air temperature control device control parameter 5F, preferably for controlling the temperature of the air 27; and / or Preferably, after the drying process, a waste conveyor driver 6G, controllable by the exhaust conveyor driver control parameters 5G, for extracting air 27, and / or A fluidized bed driver 6H, controllable by fluidized bed driver control parameters 5H, preferably for rotating the carousel of the dryer 15.

[0141] The sensors 8 for measuring the respective corresponding state parameters 7 of the granulator 13 include one or more of the following sensors 8: a feed conveyor sensor 8A for measuring a feed conveyor parameter 7A characterizing the feed, in particular the throughput and / or speed of the feed, preferably the conveying speed or measurement expressed in [kg / h]; and / or a granulator sensor 8B for measuring granulator parameters 7B characterizing the functional properties of the granulator 13, in particular the torque and / or speed of its driver 6B, and / or the throughput, preferably the conveying speed of the granulator expressed in [kg / h] or the corresponding rotation speed of the extruder / screw [rpm], and / or an injection sensor 8C for measuring injection device parameters 7C characterizing the addition of liquid 13B, in particular water and / or alcohol, to the raw material 3 in the region of the granulator 13, in particular the throughput and / or the quantity ratio compared to the raw material, preferably the spray rate expressed in [g / min], and / or One or more granulator temperature sensors 8D (preferably in units of [°C] etc.) for measuring one or more granulator temperatures 7D, in particular the temperature of the granulator 13 or (indirectly) the raw material 3 at different positions along the conveying path of the raw material 3 through the granulator 13.

[0142] The sensors 8 for measuring the respective corresponding state parameters 7 of the dryer 15 include one or more of the following sensors 8: The (mass) throughput, (corresponding) pressure difference, and / or velocity (preferably [m 3 a supply air conveyor sensor 8E for measuring a supply air conveyor parameter 7E characterizing the supply air conveyor parameter (expressed as [times] / hour); and / or a supply air sensor 8F for measuring supply air parameters 7F characterizing the properties of the supply air 27, in particular the temperature (preferably expressed in [°C]) and / or moisture content (preferably expressed in %, mass % or dew point [°C]) of the supply air 27, and / or an exhaust sensor 8G for measuring exhaust parameters 7G characterizing the exhaust air transport of the dryer 15, in particular the (mass) throughput, (corresponding) pressure difference, temperature and / or moisture of the air 27 emitted from the dryer 15; and / or A fluidized bed sensor 8H for measuring a fluidized bed parameter 7H, preferably characterizing a drying-related variable characteristic associated with the fluidized bed of the dryer 15, in particular the speed, such as the carousel speed (preferably expressed in [rpm]) of the carousel of the dryer.

[0143] Alternatively or additionally to the use of sensors 8, in principle, the characteristics or control parameters 5 of actuators 6 may also be partly used to determine or derive one or more of the aforementioned or corresponding quantities, which may then be used as a basis for control, as appropriate.

[0144] In contrast, the determination of state parameters 7, which are influenced by the raw materials, requires measurements by sensors 8. These state parameters 7 are also used as unit variables in the model 12.

[0145] Optionally, as part of the installation 1, alternatively or additionally, the following are provided externally: a formulation temperature sensor 8I for measuring a formulation temperature 7I characterizing the temperature of the formulation 2; a formulation outlet quantity sensor 8J for measuring a production and / or installation 1 characteristic relating to the production of formulation 2, in particular a formulation outlet quantity parameter 7J characterizing the outlet quantity of formulation 2, and / or A formulation property sensor 8K that characterizes an inline measurable formulation property parameter 7K of formulation 2, in particular particle size, particle size distribution, particle size, and / or moisture, particularly preferably particle size or particle size distribution (XD10, XD50, XD90), and / or residual moisture and / or drying loss of the intermediate product 14 during processing into formulation 2.

[0146] The aforementioned control parameters 5, actuators 6, state parameters 7, and sensors 8 are particularly preferred examples for an exemplary embodiment of a particularly preferred combination of a granulator 13 (preferably a (twin) screw granulator 13) and a dryer 15 (particularly preferably a fluidized bed dryer 15).

[0147] It is understood that other control parameters 5, actuators 6, state parameters 7, and sensors 8 are alternatively or additionally possible in other granulators 13 and / or dryers 15. Therefore, the present invention is preferably not limited to the control parameters 5, actuators 6, state parameters 7, and sensors 8 described above. Furthermore, in the following cases, it is not necessary to implement or use all control parameters 5, actuators 6, state parameters 7, and sensors 8. Various choices are possible here.

[0148] The production of formulation 2 by means of installation 1 will be described in more detail below with reference to the exemplary embodiment shown in Figure 1. It is understood that the present invention can in principle be particularly preferably and advantageously implemented by means of installation 1 described below, but is not limited thereto. In particular, it is not necessary to implement all of the described components of installation 1 or all of the steps performed using installation 1.

[0149] For example, it is conceivable to use the present invention when other granulator and dryer technologies are used. Thus, a granulator of a different type than the twin-screw granulator 13 can be used. Alternatively or additionally, a dryer of a different type than the fluidized bed dryer 15 can be used. Nevertheless, the present invention proves to be particularly advantageous in this context.

[0150] In the exemplary embodiment shown in FIG. 1, facility 1 includes a granulator 13 for processing raw materials 3 to form intermediate product 14, and a dryer 15 for forming formulation 2 from intermediate product 14.

[0151] The granulator 13 has a feed conveyor 16 for feeding and / or metering the raw material 3. The feed conveyor 16 may have, in particular, a funnel-shaped storage container 16A for holding the raw material 3. Furthermore, the feed conveyor 16 may have a feeding device 16B for feeding the raw material 3 at a specific feeding rate (amount per time) to a granulation unit 17 of the granulator 13 coupled to the feed conveyor 16.

[0152] The feed conveyor 16 may have a driver 6A for the feeding device 16B, in particular a motor for driving a worm or screw 16C. In general, however, other principles of feeding than by means of a screw 16C are also possible, such as a conveyor belt. The driver 6A is preferably controllable by the feed conveyor parameters 7A, as described above. The feed conveyor driver control parameters 5A can preferably be used to set the feeding speed.

[0153] The feed conveyor 16 preferably has one or more sensors 8A for measuring throughput and / or (preferably corresponding to) speed.

[0154] The feed conveyor 16 is preferably followed by a granulation unit 17 of the granulator 13. The granulation unit 17 is configured to modify the granularity and / or particle size distribution of the raw material 3. For this purpose, the granulation unit 17 can physically act on the raw material 3, in particular by kneading and / or milling / tumbling.

[0155] In the illustrated embodiment, the granulation unit 17 has at least one screw 18, preferably a twin screw. The twin screws 18 preferably intermesh and convey the feedstock 3 while physically acting on the feedstock 3 to modify the granularity and / or particle size distribution of the feedstock 3.

[0156] The granulation unit 17 and / or the screw 18 may have different processing zones 19. In particular, different screw flight pitches and / or surfaces may be provided to achieve the desired processing of the raw material 3.

[0157] The granulation unit 17 may have a granulator driver 6B for driving the granulation unit 17, in particular the screw 18. As already explained, the granulator driver 6B is controllable via the granulator driver control parameters 5B, preferably with respect to throughput and / or speed, in particular the rotational speed of the screw 18.

[0158] The granulator 13 preferably has a granulator sensor 8B which can measure parameters representative of the granulation speed and / or processing intensity, in particular the speed, the throughput or, very particularly preferably, the torque (of the granulation unit 17 / of the screw 18).

[0159] The granulator 13 may have an injection device 6C for injecting a liquid 13B for the purpose of mixing and / or blending with the raw material 3. This may be a sprayer, but alternatively a dripper or generally a device for adding liquid substances.

[0160] The injection device 6C is preferably arranged in the region of the inlet 13A and / or in the first half and / or first third of the transport path formed by the granulation unit 17 for the raw material 3 between the inlet 13A and the intermediate product outlet 21 of the granulator 13 for the raw material 3 processed in the granulator 13.

[0161] The granulator 13 preferably has at least one, but preferably more, granulator temperature sensors 8D for measuring the temperature 7D of the raw material 3 or the corresponding temperatures 7D of the granulator 15 or its housing 22 at various locations.

[0162] In the illustrated embodiment, more than three and / or less than ten granulator temperature sensors 8D are provided for measuring the temperature 7D of the raw material 3 or the corresponding temperature 7D of the granulator 15. The temperature sensors 8D are preferably provided distributed (at least substantially equidistantly) along the transport path of the raw material 3 in the granulator 15 and / or granulation unit 17.

[0163] The coupling device 20 allows for continuous and / or uninterrupted movement of the intermediate product 14 from an intermediate product outlet 21, preferably formed by a housing 22 of the granulator 13, to an intermediate product inlet 24 of the dryer 15, preferably formed by a housing 23 of the dryer 15.

[0164] Preferably, the intermediate product 14 is not stopped for the purpose of intermediate storage for at least 1 minute, 2 minutes, or more than 5 minutes. Thus, in particular, there is no sampling and analysis of the intermediate product 14 separately from the installation and / or no delay or waiting of the production process. Rather, in the sense of the present invention, it is preferred that the intermediate product 14 is fed to the dryer 14 at least substantially uninterrupted and / or continuously.

[0165] One or more of the granulator temperature sensors 8D can measure the temperature 7D of the intermediate product 14 or a corresponding temperature 7D, in particular the temperature 7D of the granulator 15 in the region of the combining device 20.

[0166] Dryer 15 preferably has a fluidized bed 25. Furthermore, dryer 15 preferably has an air inlet 26 for the inlet of (processing) air 27, an optional diffuser 28 for uniformly filling fluidized bed 25 with air 27, and an air outlet 29 for discharging air 27 after passing through fluidized bed 25. In the region of or before air outlet 29, dryer 15 optionally has a separator 30, such as a cyclone separator or a filter, for capturing particles of intermediate product 14 and / or formulation 2 from air 27.

[0167] Finally, the dryer 15 preferably has a formulation outlet 31 for discharging the formulation 2, i.e., the intermediate product 14 dried in the dryer 15. The dryer 15 dries the intermediate product 14 entering through the intermediate product inlet 24 with air 27 and subsequently discharges the formulation 2 formed thereby from the formulation outlet 31, i.e., preferably after passing through the fluidized bed 25.

[0168] To supply air 27 to the dryer 15, i.e., in particular to the fluidized bed 25, the dryer 15 preferably has a supply air conveyor 15A equipped with a supply air conveyor driver 6E, which may be a fan or generally a device for moving and / or compressing the air 27.

[0169] The dryer 15 may have a supply air conveyor sensor 8E capable of measuring the throughput or a corresponding quantity of air 27. In particular, the supply air conveyor sensor 8E is a pressure sensor for measuring the air pressure on the discharge side and / or on the side of the supply air conveyor driver 6E facing the fluidized bed 25 and / or a differential pressure sensor for determining the differential pressure across the supply air conveyor 15A. Alternatively or additionally, the supply air conveyor sensor 8E may be or have a quantity assigned to the supply air conveyor driver 6E, such as rotational speed (fan rotation speed), current consumption, torque, etc.

[0170] The supply air conveyor 15A and / or its supply air conveyor driver 6E can be controlled by supply air conveyor driver control parameters 5E, in particular with respect to throughput, (fan) rotation speed, current consumption, and / or pressure or differential pressure. The differential pressure can be the pressure across the supply air conveyor 15A, but alternatively or additionally it can be the pressure across the fluidized bed 25, etc.

[0171] The air 27 is preferably temperature-controlled, in particular heated, before being fed to the fluidized bed 25 or another drying device of the dryer 15. For this purpose, the dryer 15 preferably has a supply air temperature control device 6F, preferably a heating register, which can be controlled by or configured for supply air temperature control device control parameters 5F.

[0172] Preferably, the temperature and / or (relative) moisture of the conditioned air 27 that is in contact or brought into contact with the intermediate product 14 for drying is measured. For this purpose, the dryer 15 can have or is configured to have a supply air sensor 8F that measures the temperature of the air 27 as a supply air parameter 7F, and alternatively or additionally the (relative) moisture.

[0173] A supply air sensor 8F may be provided between the fluidized bed 25 and the supply air temperature control device 6F or the supply air conveyor 15A and / or may measure the properties of the air 27.

[0174] After air 27 has been brought into contact with intermediate product 14 for drying, air 27 is discharged by dryer 15. This is preferably done via air outlet 29 which is different from formulation outlet 31 for discharging formulation 3.

[0175] The air 27 can be delivered from the exhaust conveyor 15B, in particular via a (second) fan, through the air outlet 29. For this purpose, an exhaust conveyor driver 6G can be provided which effects the transport and / or drives the exhaust conveyor 15B.

[0176] The exhaust conveyor 15B can be open-loop or closed-loop controlled by the exhaust conveyor driver control parameters 5G. In particular, the exhaust conveyor 15B is closed-loop controlled so that no air 27 leaks through the intermediate product outlet 21 and the formulation inlet 31. To this end, the conveying rate of the exhaust conveyor 15B can be equal to or exceed the conveying rate of the supply air conveyor 15A.

[0177] The exhaust conveyor sensor 8G can be used to measure the throughput, velocity, temperature, and / or moisture of the air 27 exhausted from the dryer 15 after the drying process, and / or corresponding quantities such as the speed of the exhaust conveyor driver 6G and / or fan blade wheel as exhaust parameters 7G.

[0178] The fluidized bed 25 may have or form a processing zone, in particular opposite the air inlet 26. The fluidized bed 25 and / or the structures that delimit it in the direction of the air inlet 26 (such as screens or perforated plates and / or carousels) may be drivable, in particular actuable. For this purpose, a fluidized bed driver 6H may be provided that is controllable by means of fluidized bed driver control parameters 5H.

[0179] Fluidized bed sensors 8H can be used to measure fluidized bed parameters 7H, which can characterize properties of the fluidized bed 25, such as the movement of the carousel.

[0180] After drying the intermediate product 14 in a dryer 15, the processed raw material 3 is output as a formulation 2.

[0181] From formulation 2, one or more actual formulation parameters 11, i.e., parameters describing the physical or chemical properties of the manufactured formulation 2, can be determined at this instance or later. This can be done in-line, i.e., in an uninterrupted continuous process. Alternatively or additionally, however, the actual formulation parameters 11 can also be determined later by laboratory investigations. The actual formulation parameters 11 that cannot be measured in-line are preferably taken into account as the basis for model 12. Thus, the modeling takes into account the actual formulation parameters 11 that cannot be measured in-line. In contrast, the actual formulation parameters 11 that cannot be measured in-line are not used directly for controlling facility 1 or as input variables for control 9 and / or model 12.

[0182] The equipment 1 may have one or more sensors 8 for in-line characterization of properties of the formulation 2. These include one or more of a formulation temperature sensor 8I for measuring the formulation temperature 7I, a formulation outlet volume sensor 8J for measuring a formulation parameter 7J describing the outlet volume and / or throughput (mass flow rate) of the formulation, and / or a formulation property sensor 8K for measuring one or more properties of the formulation 2 and outputting, as formulation property parameter 7K, preferably (relative) moisture and / or residual moisture and / or loss on drying.

[0183] Formulation parameter 10, 11 is preferably at least one parameter characterizing the particles of formulation 2, such as particle size or particle size distribution (XD10, XD50, and / or XD90) or a corresponding amount thereof.

[0184] Alternatively or additionally, the formulation parameters 10, 11 are preferably (relative) moisture and / or residual moisture and / or loss on drying (LoD - loss on drying) and / or a corresponding amount thereof.

[0185] Additional formulation parameters 10, 11 can be determined as needed by formulation temperature sensor 8I, formulation outlet volume sensor 8J, and / or formulation characteristic sensor 8K, and / or as formulation temperature 7I, formulation outlet volume parameter 7J, and / or formulation characteristic parameter 7K.

[0186] In principle, corresponding measured variables such as particle size or particle size distribution (XD10, XD50, and / or XD90) of formulation 2 can be used to form model 12. However, in any case it is not necessary or provided to determine corresponding quantities in the ongoing operation of installation 1 or to supply said quantities to its control.

[0187] Optionally, but preferably, an in-line measurable property of the intermediate product 14, in particular moisture, can be determined from the intermediate product 14 by means of an intermediate product (moisture) sensor 8L as an intermediate product property parameter 8L. If provided, this can also be taken into account in the model 12, in particular used as an input variable and / or (additionally) used as the basis for controlling the installation 1.

[0188] Actual formulation parameters 11 can be measured downstream of facility 1. This will be discussed in more detail below. However, it should already be mentioned at this point that parameters characterizing the shape, size or distribution of the particles of formulation 2 are also preferably measured in-line.

[0189] In contrast, measurements that characterize the particles of intermediate product 14 are preferably avoided.

[0190] In connection with the exemplary embodiment according to Fig. 1, various sensors 8 have been described for determining the state parameters 7 of the installation 1. However, it is not necessary that all sensors 8 are provided and / or that all state parameters 7 are used.

[0191] Preferably, at least two or at least three state parameters 7 and / or sensors 8 are used in the granulator 13 and dryer 15, respectively.

[0192] Control parameters 5 for the equipment control (which can be particularly preferably dynamically set for the equipment control, the selection of which can represent an independent concept of the invention for further aspects of the invention) include: Regarding Granulator 13, Feed conveyor driver control parameters 5A, preferably characterizing the metering of the raw material 3, and / or Granulator driver control parameters 5B, which preferably characterize the extruder speed of the extruder of the granulator 13, which may form the granulation unit 17 of the granulator 13, or (other) quantities corresponding to the conveying speed and / or processing speed of the granulator 13, and / or In particular, the injection device control parameter 5C is the spray rate, which characterizes the quantity per hour of the liquid 13B dispensed.

[0193] Regarding dryer 15, a fluidized bed driver control parameter 5H, which preferably characterizes the speed of the carousel of the dryer 15; Supply air conveyor driver control parameter 5E, preferably representing the inlet supply air flow, and / or In particular, the supply air temperature control device control parameter 5F represents the temperature of the supply stream on the inlet side of the air 27.

[0194] Parameters 3, 7 that are particularly preferred as input variables for the plant control and / or model (the selection of which may represent an independent concept of the invention for further aspects of the invention) include: Regarding Granulator 13, granulator parameters 7B, which characterize in particular the extruder torque of the extruder of the granulator 13 or other parameters of the granulator driver 6B, which depend on the consistency and / or the conveying speed of the raw material 3 to be processed, and / or one or more granulator temperatures 7D characterizing one or more temperatures of the raw material 3 being processed in the granulator 13, or temperatures corresponding thereto, in particular at different positions along the material flow of the raw material 3 in the granulator 13, and / or Granulator temperature 7D and / or intermediate product characteristic parameter 7L which is or corresponds to the temperature of the intermediate product.

[0195] Regarding dryer 15, Exhaust parameters 7G characterizing in particular the exhaust temperature and / or (relative) moisture and / or throughput (e.g. expressed by a pressure difference) of the exhausted air 27, and / or In particular, a product outlet volume parameter 7J characterizing the outlet volume and / or pressure differential associated with the output of the product, for example, through a filter or screen.

[0196] Regarding ingredient 3, A raw material parameter 4 preferably characterizing the moisture and / or particle size of the raw material 3 .

[0197] Optionally, for formulation 2, the following can additionally be taken into account for the control of equipment 1: Preferably formulation characteristic parameters 7K characterizing particle size distribution, in particular D10, D50 and / or D90, and / or loss on drying.

[0198] The use of parameters 4, 5, 7 above may be supplemented by one or more of parameters 4, 5, 7 discussed above and below.

[0199] FIG. 3 shows a simplified schematic diagram of an artificial neural network 32 for determining control parameters 5 for controlling the plant 1 and / or its actuators 6.

[0200] According to one aspect of the present invention, the model 12, and in particular the static part 12A of the model 12, is formed by or includes an artificial neural network 32. An example structure of the artificial neural network 32 is shown in FIG.

[0201] The artificial neural network 32 includes nodes 36 in the input layer 33 in the form of input nodes corresponding to one or more of the raw material parameters 4, one or more of the state parameters 7 of the granulator 13, one or more of the state parameters 7 of the dryer 15, and / or one or more of the target formulation parameters 10.

[0202] The artificial neural network 32 preferably includes nodes 36 in one or more hidden layers 34 through which the input layer 23 can be linked to the output layer 35 .

[0203] The artificial neural network 32 may include nodes 36 in the form of output nodes in an output layer 35 that correspond to one or more of the control parameters 5 .

[0204] Nodes 36 in different layers 33, 34, 35 may be connected to each other by edges 37. The nodes 36 may form a graph by the edges 37. The nodes 36 and / or edges 37 preferably have weights 38, which specify the link of nodes 36 represented by the edges 37.

[0205] The artificial neural network 32 is preferably trained with a data set consisting of different combinations of raw material parameters 4, state parameters 7, and control parameters 5, and the actual formulation parameters 11 that result when these parameters are given.

[0206] The training data set represents a steady state of the equipment 1 in which the actual formulation parameters 11 and the state parameters 7 assume at least substantially static values ​​based on the constant control parameters 5 and the raw material parameters 4, respectively.

[0207] The artificial neural network 32 is preferably trained by applying to the input nodes 36 at least one, and preferably more, preferably feedstock-influenced state parameter 7 of the granulator 13 and at least one, and preferably more, corresponding, preferably feedstock-influenced state parameter 7 of the dryer 15 for each training data set.

[0208] The input nodes (nodes 36 in input layer 33) are preferably further each applied with one or more corresponding real formulation parameters 11 and one or more corresponding ingredient parameters 4 of a respective training data set.

[0209] Specifying the parameters 4, 5, 7, 11 of the training data set in the input layer 33 provides control parameters 5 (values ​​of the neural network 32) to the output nodes (nodes 36 in the output layer 35). Errors are preferably determined from the control parameters 5 by comparison with the control parameters 5 of the respective training data set, and the errors are reduced and / or compensated for, preferably by backpropagation and / or by continuously adjusting the weights 38 of the artificial neural network 32.

[0210] Next, instead of the actual formulation parameters 11, the target formulation parameters 10 are specified / provided to the artificial neural network 32 (in the input layer 33) together with further current parameters 4, 7 in order to finally determine the control parameters 5 using the neural network 32, and then the control parameters 5 for controlling the equipment 1 are obtained (in the output layer 35), based on which the equipment 1 can be controlled or is (automatically) controlled.

[0211] The parameters 4, 5, 7, 10 to which the node 36 is provided are preferably at least the following: Preferably, in the input layer 33: at least one node 36 for at least one corresponding feed parameter 4, preferably the moisture content of the feed 3 and / or a property of the particles of the feed 3, in particular their particle size distribution; node 36 for the granulator parameters 7B, in particular the screw torque and / or extruder torque of the granulator 13, and / or a node 36 for a granulator temperature 7D, in particular a plurality of nodes 36 for a plurality of granulator temperatures 7D, and / or a node 36 for exhaust parameters 7G, ​​in particular the temperature and moisture content of the air 27 and / or the pressure difference of the air 27 across the separator 30; Preferably, in the output layer 35: node 36 for granulator driver control parameters 5B, and / or Node 36 for infusion device control parameters 5C, and / or Node 36 for fluidized bed driver control parameters 5H, and / or • Node 36 for supply air conveyor driver control parameters 5E and / or node 36 for supply air temperature control device control parameters 5F.

[0212] Optionally, nodes for one or more of the following state parameters 7 are provided in the input layer 34: Node 36 for injection device parameters 7C; Node 36 for supply air parameter 7F, Node 36 for supply air conveyor parameter 7E, ● Node 36 for fluidized bed parameter 7H, ● Node 36 for formulation temperature 7I, Node 36 for formulation exit volume parameter 7J, and / or ● Node 36 for formulation characteristic parameter 7K; and / or one or more of the following control parameters 5 in the output layer 35 are provided: Node 36 for feed conveyor driver control parameters 5A; Node 36 for granulator temperature control device control parameters 5D, and / or ●Node 36 for exhaust conveyor driver control parameter 5G.

[0213] Therefore, in each case, the state parameters 7 to which node 36 corresponds preferably include one or more of granulator parameters 7B, in particular the torque of the granulation unit of granulator 13, granulator temperatures 7D at different positions along the transport path of granulator 13 for raw materials 3, formulation temperatures 7I, in particular the temperature of formulation 2 at formulation outlet 31 of dryer 15, formulation characteristic parameters 7K, in particular the moisture, in particular the relative moisture, of formulation 2 at formulation outlet 31 of dryer 15, and / or exhaust temperature 7G, in particular characterizing the pressure drop across the filter of dryer 15, here, for example, (cyclone) separator 30.

[0214] The artificial neural network 32 is preferably configured by training to generate, from the parameters 4, 7, 10 supplied to the nodes 36 of the input layer, control parameters 5 capable of controlling the plant 1 and / or the combination of the granulator 13 and the dryer 15. To this end, the control parameters 5 generated by the artificial neural network 32 are used as the basis for controlling the plant 1, preferably, but not necessarily, after being optimized by the dynamic model 12B.

[0215] Using the model 12, the control parameters 5 are preferably determined based solely or primarily on the raw material parameters 4, the state parameters 7, and the target formulation parameters 10.

[0216] Again, properties of the intermediate product 14 that cannot be measured in-line preferably remain unconsidered. Preferably, at most the temperature and / or moisture of the intermediate product 14 are considered.

[0217] The model 12 preferably takes into account, by prediction, the future effects of changes in the state parameters 7 on the actual formulation parameters 11. To this end, the dynamic part of the model 12 is preferably configured to take into account changes in the actual formulation parameters 11 caused by long-term effects.

[0218] The control 9 preferably takes into account the future impact of changes in the control parameters 5 on the actual formulation parameters 11 and / or state parameters 7, preferably by prediction. In particular, the dynamic part of the model 12 is configured to proactively compensate for changes in the actual formulation parameters 11 caused by long-term effects.

[0219] For this purpose, the model 12 can use predictions about the future development of the actual formulation parameters 11 as a basis for determining or adjusting the control parameters 5. For this purpose, the model 12 can have a static part 12A in which the base values ​​of the respective control parameters 5 are determined from the state parameters 7 and, preferably, from the raw material parameters 4 and the target formulation parameters 10, as well as a dynamic part 12B in which the base values ​​are optimized by the predictions.

[0220] Therefore, using the dynamic part 12B of the model 12, it is possible to predict the change in the actual formulation parameter 11 when the current state parameter 7 changes based on the raw material parameter 4 and the current state parameter 7, and based on this prediction, the basic value of the control parameter 5 can be adjusted and the equipment 1 can be controlled with the adjusted control parameter 5.

[0221] Thus, model 12, and in particular dynamic portion 12B of model 12, is configured to predict the long-term effects of changes in control parameters 5 on actual formulation parameters 11 and / or state parameters 7. Control 9 is therefore configured by model 12 to control equipment 1 and / or the combination of granulator 13 and dryer 15 while compensating for the long-term effects. Surprisingly, therefore, and despite the direct coupling of granulator 13 and dryer 15, production can be carried out while maintaining small / acceptable deviations from target formulation parameters 10 to actual formulation parameters 11.

[0222] If only prediction of the properties of formulation 2 is desired or achieved using model 12, model 12 and / or artificial neural network 32 can be constructed differently than in the case of preferably fully automatic control by model 12, so as to be able to determine and / or output properties characterizing formulation 2, and therefore in particular one or more predicted actual formulation parameters 11.

[0223] In this case, using the model 12, one or more actual formulation parameters 11 are determined and / or predicted, preferably based solely or primarily on the raw material parameters 4, the state parameters 7, and the specified / preset control parameters 5.

[0224] Thus, the dynamic portion 12B of the model 12 can be used to predict changes in the actual formulation parameters 11, preferably based on the raw material parameters 4, the current state parameters 7, and / or the control parameters 5, and / or their base values, taking into account changes in the state parameters 7 as the control parameters 5 change, and based on this prediction, the resulting actual formulation parameters 11 can be predicted and preferably output.

[0225] Thus, model 12, and in particular dynamic portion 12B of model 12, is preferably configured to predict the long-term effect of changes in state parameters 7 on actual formulation parameters 11. As a result, and surprisingly, and despite the direct coupling of granulator 13 and dryer 15, the predicted actual formulation parameters 11 can be used to provide guidance to the user for selecting appropriate control parameters 5.

[0226] Regardless of whether the control parameters 5 or the predicted actual formulation parameters 11 are determined by the model 12, properties of the intermediate product 14 that cannot be measured in-line, such as, for example, physical properties of the intermediate product 14 that characterize the particles of the intermediate product 14, preferably remain unaccounted for. In particular, all properties of the intermediate product 14 remain unaccounted for, except for the temperature and moisture of the intermediate product 14.

[0227] Thus, the model 12 preferably takes into account at most these parameters of the feedstock 2 and / or intermediate product 14 being processed and can be measured in-line, thereby avoiding the need for sampling and analysis separate from the facility.

[0228] As already mentioned, model 12 may alternatively be configured to predict actual formulation parameters 11. In this case, model 12 and / or artificial neural network 32 may be configured differently from the artificial neural network 32 for determining control parameters 5.

[0229] When using the model 12 to determine predicted actual formulation parameters 11, the parameters 4, 5, 7, 10 for which nodes 36 are provided are preferably at least the following: Preferably, in the input layer 33: a node 36 for at least one corresponding feedstock parameter 4, preferably the moisture content of the feedstock 3, and / or a property of the particles of the feedstock 3, in particular their particle size distribution, and / or node 36 for the granulator parameters 7B, in particular the screw torque and / or extruder torque of the granulator 13, and / or a node 36 for a granulator temperature 7D, in particular a plurality of nodes 36 for a plurality of granulator temperatures 7D, and / or a node 36 for exhaust parameters 7G, ​​in particular the temperature and moisture of the air 27 and / or the pressure difference of the air 27 across the separator 30; and / or node 36 for granulator driver control parameters 5B, and / or Node 36 for infusion device control parameters 5C, and / or Node 36 for fluidized bed driver control parameters 5H, and / or Node 36 for supply air conveyor driver control parameters 5E, and / or ● Node 36 of supply air temperature control device control parameter 5F. Preferably, in the output layer 35: ● One or more nodes 36 (respectively) for the (predicted) actual formulation parameters 11 .

[0230] Optionally, nodes 36 for one or more of the following state parameters 7 are provided in the input layer 34: Node 36 for injection device parameters 7C; Node 36 for supply air parameter 7F, Node 36 for supply air conveyor parameter 7E, ● Node 36 for fluidized bed parameter 7H, ● Node 36 for formulation temperature 7I, ● Node 36 for formulation parameter 7J, ● Node 36 for formulation characteristic parameter 7K; Node 36 for feed conveyor driver control parameters 5A, and / or Node 36 for granulator temperature control device control parameters 5D, and / or ●Node 36 for exhaust conveyor driver control parameter 5G.

[0231] In this case, the artificial neural network 32 may be trained on a corresponding and / or the same data set as the artificial neural network 32 for determining the control parameters 5 .

[0232] FIG. 4 shows a schematic diagram of the past and predicted course of one or more parameters 5, 7, 10, and 11.

[0233] To the left of the Y-axis, which represents the values ​​of parameters 5, 7, 11, is the past evolution of one or more parameters 5, 7, 11, as indicated by the arrow labeled Past P, and the past course of the reference trajectory 39 and the measured parameters 7, 11. Additionally, their future evolution is shown in the forecast period 40, as indicated by the arrow labeled Future F.

[0234] According to the proposal, one or more of parameters 5, 7, and 10 are discrete time t k+p In an exemplary embodiment, the time t k+p are separated from one another by a sampling / scanning time Δt. In principle, however, it is not necessary that the sampling time Δt be constant, even if possible, and the sampling time Δt can be selected to be small, so that the course of one or more parameters 5, 7, 11 can be at least substantially continuous.

[0235] As shown in FIG. 4, as a result of the prediction of one or more parameters 5, 7, 11, different courses, both ascending and descending courses, can occur with the aim of approximating the measured values, such as the state parameter 7 and / or the actual formulation parameter 11, to the reference trajectory 39.

[0236] Finally, the progression according to FIG. 4 represents a possible system behavior that allows for an advantageously optimized control of the installation 1 and / or the manufacturing process by the model 12, together with the past evolution and the evolution of various quantities that are expected to develop in the future that are taken into account.

[0237] 5 shows a schematic diagram of the results of control with constant control parameters 5 over time. One or more state parameters 7 are obtained from the at least substantially constant control parameters 5 in a pre-set, preferably fixed, time window 41. Here, the state parameters 7 asymptotically approach fixed values, taking into account quality dynamics 42, which can preferably be measured in discrete time.

[0238] Based on certain control parameters 5 or a combination of certain control parameters 5 and the state parameters 7 derived therefrom, the static part of the model 12A can be determined. In particular, based on this, a machine learning-based model 12A of the static equipment behavior can be generated. This can be an artificial neural network 32, as mentioned above, although in principle other machine learning-based methods are also possible.

[0239] Figure 6 shows a schematic diagram of mutually temporally offset predictions over time t. k+p In , future measurements k depend on previous evolutions and current control parameters 5 and / or state parameters 7. Such regressions can be solved by time series forecasting, as shown in Figure 6. It shows the progression of measurements k at different relative times ζ, which are offset in time from each other by a time lag and / or sampling time Δt, thereby indicating that they are forecasts of a time series.

[0240] In particular, it is possible, and preferably, for the predictions made by the model 12 to be updated at regular intervals Δt. When the conditions of the installation 1 change, and thus one or more state parameters 7 deviate from their previous values, it is preferably possible, using the model 12, to generate modified and / or adjusted predictions, in particular of one or more control parameters 5, state parameters 7, and / or actual formulation parameters 11.

[0241] Figure 7 shows a schematic diagram of the process characteristics over time. The basic idea is to combine the static behavior, as explained for example with reference to Figure 5, with the dynamic behavior and predicted evolution taking into account past events, so that, as shown by the example of Figure 7, a process with static base values, and therefore preferably determined by the static part 12A of the model 12, is started, and from then on, during production, by repeated iterative optimization, in particular by a time-prediction approach, the desired attributes, in particular therefore the actual formulation parameters 11, can be achieved in a short time and then at least substantially maintained.

[0242] Thus, in principle, the control parameters 5 can be determined based on the static portion 12A of the model 12 at the start of continuous processing of the raw material 3 to form the formulation 2, and the control parameters 5 can be easily readjusted / updated during ongoing operation of the equipment 1 by the dynamic portion 12B of the model 12. This is shown in time in the time section of Figure 7, in the area where the sampling time Δt is plotted.

[0243] Thus, the installation 1 first reaches a quasi-steady state before readjustment is carried out by the dynamic part 12B of the model 12, which is then preferably set by time series forecasting at intervals of each sampling time Δt. However, in principle, other control strategies are also possible.

[0244] The proposed installation 1 can be used particularly advantageously in a system 45 for the production of tablets. An extended method for this purpose is shown in Figure 8, with reference to the schematic flow diagram. The system 45 can partially provide components for processing the raw materials 3 and / or components for post-processing the formulation 2, such as preferably a pneumatic conveyor installation 46, which adds additional functions to the installation 1, as will be explained in more detail below with reference to Figure 9.

[0245] Figure 9 shows an installation 1 embedded in a system 45. The preparation of raw material 3 can be upstream of installation 1. In the exemplary embodiment according to Figure 9, and with reference to the method according to Figure 8, raw material 3 is produced from components 47 of raw material 3 by sieving and / or mixing in a preparation step 48. In this case, by way of example and generally preferably, a powder mixture of components 47 is produced.

[0246] In a granulation step 49, preferably carried out continuously according to the proposal with a subsequent drying step 50, the raw material 3 is subsequently processed by a granulator 13 to form an intermediate product 14, preferably formed by adding a liquid 13B (also called granulation liquid). In an ongoing continuous process, the intermediate product 14 is then dried by a dryer 15, thereby finally producing the formulation 2. With regard to the continuous processing of the raw material 3 to form the formulation 2, reference is additionally made to the previous section.

[0247] Formulation 2 can then be further processed. In one or more post-processing steps 51, formulation 2 can be sieved, for example in a post-processing device 52, particularly for particle selection, to form a post-processed (particularly sieved) formulation 53. Alternatively or additionally, formulation 2 and / or post-processed formulation 53 are mixed to form a final mixture 54 in a mixing process 55 with additives such as disintegrants and / or binders 56.

[0248] Finally, dosage form 58, in particular one or more tablets, can be produced from formulation 2 and / or post-processed formulation 53 or final blend 54 by a tableting process 57, in particular compression.

[0249] The proposed, preferably formulation process focuses on sequential process steps for the manufacture of solid oral dosage forms, as will be described by way of example with additional reference to the schematic flow diagram of FIG. 8 and system 45 from FIG. 9.

[0250] After the optional, preferably batchwise, preparation of a homogeneous powder premix of components 47 as raw material 3, preferably by sieving and / or mixing of powders to form raw material 3 as components 47, intermediate product 14, preferably a wet granulate (wet or moist granulate), is formed in a continuous granulation step using granulator 13, preferably a so-called twin-screw granulation (TSG). For this purpose, liquid 13B can be added to raw material 3 and / or during the granulation process in granulator 13. This can be done using injection device 6C.

[0251] The intermediate product 14 is preferably conveyed in a continuous product stream directly to a continuously operating dryer 15, here a fluidized bed dryer. After two continuous process steps of granulation and drying, Formulation 2, preferably a dry granulation (dried wet granulation), is obtained.

[0252] Formulation 2 is optionally and preferably then sieved (to form post-processed formulation 53) and / or mixed with an extragranular phase (excipient / disintegrant and / or binder 56) to obtain a final blend 54. This final blend 54 (target blend) preferably forms the starting material for the tableting process 57 and is used for tableting, or system 45 is designed for this purpose.

[0253] Thus, preferably two continuous process steps, granulation and drying, are provided as equipment 1. In contrast, the system preferably combines equipments 1 to form a total of at least three, in particular four, different process units and / or production steps.

[0254] The continuous granulation process involves the process units of filling and twin-screw wet granulation. The continuous drying process involves two process units: continuous fluidized bed drying and a pneumatic conveyor system.

[0255] In total, preferably six control parameters 5 (main input variables) are defined for and / or used to control the control / control device 9 of the continuous granulation and drying installation 1 .

[0256] The at least two, preferably at least three, control parameters 5 and / or main input variables of the granulator 13 are preferably or include the injection / metered amount [kg / h] (of raw material 3), the extruder rotation speed [rpm] (of the granulator 13), and / or the spray rate [g / min] (of the injection device 6C).

[0257] For the dryer 15, at least two, preferably at least three, control parameters 5 and / or primary input variables are the supply air flow rate [m 3 / hour], supply air temperature [°C] (of the supply air entering the dryer 15), and / or carousel rotation speed [rpm] (of the dryer 15 and / or its carousel).

[0258] In total, preferably at least six primary input and / or control parameters 5 are defined for or used to control the equipment 1, so it can be readily appreciated that understanding, controlling, and monitoring the relationships of such multi-functional equipment 1 can present challenges.

[0259] 8 and 9 show a schematic diagram of the material and data flows of the continuous granulation and drying installation 1 and / or the system 45 formed therewith.

[0260] With regard to material flow, it is preferred that system 45 operates according to a bin-to-bin approach and / or is configured for this purpose, meaning that a first vessel feeds a preferably homogenous powder premix as raw material 3 into a continuous line (from granulator 13 and dryer 15) to obtain a dry granulate as formulation 2 in a second vessel after two successively performed process steps (preferably twin-screw wet granulation and fluidized bed drying).

[0261] A powder premix as raw material 3 is preferably prepared in batches by system 45. Further processing of the dry granulation (formulation 2) is also preferably performed in batches as shown in the example of FIG.

[0262] In contrast to fully continuous manufacturing (from raw materials to finished tablets), this bin-to-bin method offers greater flexibility due to the modular nature of the system 45. Thus, one advantage is that by embedding a continuous method of granulation and drying in a bin-to-bin method, the aforementioned benefits can be achieved with greater flexibility at the same time.

[0263] Preferably, three types of data are taken into account for controlling the installation 1 and / or the system 45:

[0264] The first is the control parameters 5, which as previously mentioned preferably consist of six related control parameters 5.

[0265] The second type of data are state parameters 7 relating to process conditions (temperature, pressure drop, torque, etc.), which can be measured via multiple sensors 8 located throughout the installation 1, preferably in online and / or real-time mode.

[0266] The third type of data preferably consists of one or more raw material parameters 4 and / or actual formulation parameters 11. These important material attributes (of the raw materials 3, intermediate products 14, and / or formulations 2) are preferably measured separately from the equipment 1, discontinuously and / or based on random samples, preferably as in-process tests, thus forming a third type of data with a time delay.

[0267] A complete data set, preferably including all three data types, forms the basis for creating one or more models 12 for use within the scope of the present invention.

[0268] In particular, two related aspects can be highlighted as essential, even though they are not essential, advantages of the continuous granulation and drying installation 1 achieved by the present invention.

[0269] For example, the multifactorial interactions of six main input variables (control parameters 5 and state parameters 7 that can be directly influenced by the control parameters 5 and / or not influenced by the raw materials 3) and the resulting output variables. The resulting state parameters 7 and / or state parameters 7 influenced by the raw materials (e.g., 10 states, measured online and in real time) and material attributes (preferably, for example, attributes characterizing the raw materials 3 and formulation 2 described by a total of four material parameters and / or actual formulation parameters 11, measured offline and with a time delay) can be defined as output variables.

[0270] Furthermore, the continuous granulation and drying installation 1 preferably combines different (successive) process steps. Thus, the process and / or control parameters 5 of a process unit influence the process state and / or state parameters 7 of the next unit and finally also the material properties and / or actual formulation parameters 11.

[0271] Considering these two aspects, it becomes clear that optimal manual control of the process presents challenges and rarely leads to positive results, and the significant advantages of the preferred machine learning-based method for providing prediction and control adjustments for this continuous process become clearly visible.

[0272] In summary, the following aspects can be implemented in the installation 1 individually or in various possible combinations: - Integrated continuous process: It is not a completely continuous process (from active substances and auxiliaries to the final product), but replaces at least two traditional batch processes with one continuous process step. Combination of a batch process and a continuous step = Integrated continuous process - the use of flat-bottom metering devices as feed conveyors 16, which allow for very precise and easily controllable mass flow rates; - Use of twin-screw wet granulation equipment as granulator 13, - Using a continuous fluidized bed dryer as dryer 15 allows for dual functionality, converting the fluidized bed dryer into a fluidized bed granulator. This modular and highly versatile system allows for dual use of the main equipment, thereby increasing facility productivity, efficiently using the facility's footprint, and reducing downtime. - Use of a special continuous dryer 15 with a slowly rotating carousel that divides the large fluidized bed chamber into several smaller chambers, thus avoiding the formation of partial batches. - Reduction of the possibility of large accumulations of wet granulate (intermediate product 14) by shortening the conveying path and avoiding valves for the wet granulate. - In-line data generation of NIR probes and / or particle size measurement systems (Intermediate Product Sensor 8L / Formulation Property Sensor 8K) for measuring material properties of the wet granulate (Intermediate Product 14) and / or dry granulate (Formulation 2).

[0273] The invention further relates to a computer program product or a computer readable storage medium comprising instructions which, when the program is executed by a computer, cause the computer to carry out the proposed method / method steps or parts thereof.

[0274] The individual aspects of the present invention can be implemented separately from one another, but also in different combinations. In particular, the aspects described in relation to system 45 can be combined or may be advantageously combined with the aspects described above in relation to the combination of granulator 13 and dryer 15. [Explanation of symbols]

[0275] 1 equipment 2. Preparation 3 Raw materials 4. Raw material parameters 5 Control parameters 5A Feed conveyor driver control parameters 5B Granulator driver control parameters 5C Infusion Device Control Parameters 5D Granulator Temperature Control Device Control Parameters 5E Supply Air Conveyor Driver Control Parameters 5F Supply air temperature control device control parameters 5G exhaust conveyor driver control parameters 5H Fluidized bed driver control parameters 6 Actuators 6A Feed conveyor driver 6B Granulator driver 6C Infusion Device 6D Granulator Temperature Control Device 6E Supply air conveyor driver 6F Supply air temperature control device 6G Exhaust Conveyor Driver 6H Fluidized Bed Driver 7 State parameters [7] State parameter value 7A Feed conveyor parameters 7B Granulator Parameters 7C Injection Device Parameters 7D Granulator temperature 7E Supply Air Conveyor Parameters 7F Supply Air Parameters 7G Exhaust parameters 7H Fluidized bed parameters 7I Formulation temperature 7J Product Exit Volume Parameters 7K formulation characteristic parameters 7L Intermediate product characteristic parameters 8 sensors 8A Feed Conveyor Sensor 8B Granulator Sensor 8C Injection Sensor 8D Granulator Temperature Sensor 8E Supply Air Conveyor Sensor 8F Supply air sensor 8G Exhaust Conveyor Sensor 8H Fluidized Bed Sensor 8I Formulation temperature sensor 8J Product outlet volume sensor 8K Pharmaceutical Property Sensor 8L intermediate product sensor 9 Control Devices 10 Targeted Formulation Parameters 10A database 11 Actual formulation parameters 12 models 12A static part 12B Dynamic part 13 Granulator 13A Granulator inlet 13B Granulation liquid 14 Intermediate products 15 Dryer 15A Supply Air Conveyor 15B Exhaust Conveyor 16 Feed conveyor 16A Storage Container 16B Supply Device 16C screw 17 Granulation unit 18 Screw 19 Processing Zone 20 Coupling Device 21 Intermediate product outlet 22 Housing 23 Housing 24 Intermediate product inlet 25 Fluidized Bed 26 Air inlet 27 Air 28 Diffuser 29 Air outlet 30 separator 31 Drug Product Exit 32 Artificial Neural Networks 33 Input layer 34 Hidden Layer 35 Output layer 36 nodes 37 Edge 38 Weight 39 Reference trajectory 40 Forecast Period 41 Fixed Time Window 42 Quality Dynamics 43 Planning Dimensions 44 Process Status 45 Systems 46 Conveyor equipment 47 ingredients 48 Preparation Steps 49 Granulation Step 50 drying steps 51 Post-processing steps 52 Post-processing devices 53 Post-processed preparations 54 Final mixture 55 Mixing Process 56 Disintegrants and / or binders 57 Tableting Process 58 Dosage Forms t time t k+p time Δt sampling time P past F. Future ζ Relative time k measurement value

Claims

1. A method for controlling an equipment (1) for producing a formulation (2) from raw materials (3), the production comprising processing the raw materials (3) using a granulator (13) and drying an intermediate product (14) produced from the raw materials (3) using the granulator (13) in a dryer (15); Based on model (12), (a) control parameters (5) of the equipment (1) are determined, which are or represent manipulated variables for controlling actuators (6) of the equipment (1), and the model (12) takes into account pre-set target formulation parameters (10) which represent desired properties of the formulation (2) produced or to be produced, in particular target grain properties and moisture, or (b) actual formulation parameters (11) representing actual properties of the formulation (2) produced or to be produced, in particular actual grain properties and moisture, are predicted, said model (12) taking into account pre-defined or pre-definable control parameters (5) which are or represent manipulated variables for controlling actuators (6) of said installation (1), State parameters (7) of the equipment (1), in particular sensor values, each representing a state of the equipment (1) that influences the production, are determined and processed by the model (12), and raw material parameters (4) representing the attributes of the raw material (3), in particular moisture and / or grain properties, are taken into account by the model (12), the dryer (15) is coupled to the granulator (13) in such a way that the intermediate product (14) is automatically conveyed from the granulator (13) into the dryer (15) without interruption, and the model (12) takes into account the combination of the continuous execution of granulation using the granulator (13) and subsequent drying using the dryer (15); and / or The method of claim 1, wherein the model (12) comprises a static part (12A) by which a base value is determined for each control parameter (5) or predicted actual formulation parameter (11), and the model (12) has a dynamic part (12B) by which the base values ​​are optimized by prediction.

2. 2. The method of claim 1, wherein the model (12), in particular the static part (12A) of the model (12), is or comprises an artificial neural network (32), the artificial neural network (32) comprising nodes (36) in an input layer (33) corresponding to at least one state parameter (7) of the granulator (13), at least one state parameter (7) of the dryer (15), and preferably the raw material parameters (4).

3. the artificial neural network (32) comprises nodes (36) in the output layer (35) corresponding to the control parameters (5), while the artificial neural network (32) comprises nodes (36) in the input layer (33) corresponding to the target formulation parameters (10); or 3. The method of claim 1, wherein the artificial neural network (32) comprises nodes (36) in the input layer (33) corresponding to the control parameters (5), while the artificial neural network (32) comprises nodes (36) in the output layer (35) corresponding to the predicted actual formulation parameters (11).

4. 4. The method according to claim 2 or 3, characterized in that the artificial neural network (32) is trained or has been trained with different training data sets, each consisting of a combination of the raw material parameters (4), the state parameters (7), and the control parameters (5) and the actual formulation parameters (11) that result when these parameters are given.

5. 5. The method of claim 4, wherein each of the training data sets represents a steady state of the equipment (1), in which the actual formulation parameters (11) and the state parameters (7) assume at least substantially static values ​​based on constant control parameters (5) and raw material parameters (4).

6. 6. The method according to claim 5, characterized in that the artificial neural network (32) is trained or has been trained by applying to the nodes (36) at least one state parameter (7) of the granulator (13) and at least one corresponding state parameter (7) of the dryer (15) of each training data set, and preferably corresponding raw material parameters (4) of each training data set.

7. 7. The method according to claim 6, characterized in that each of the nodes (36) further has corresponding control parameters (5) or actual formulation parameters (11) applied to it, as long as the node (36) is provided for this purpose in the input layer (33).

8. 8. The method according to claim 6 or 7, characterized in that for each of the control parameters (5) or predicted real formulation parameters (11), values ​​are provided at the nodes (36) of the output layer (35), these values ​​are compared with the corresponding control parameters (5) or real formulation parameters (11) of each training data set to determine an error, and the error is reduced by backpropagation and / or by adjusting the weights of the artificial neural network (32) successively from training data set to training data set.

9. 9. The method according to claim 1, wherein the control parameters (5) include control parameters (5B, 5C, 5D) for controlling the granulator (13) and at least one control parameter (5E, 5F, 5G, 5H) for controlling the dryer (15).

10. 10. The method according to any one of claims 1 to 9, characterized in that the state parameters (7) comprise at least one parameter (7B, 7C, 7D) describing the operating state of the granulator (13) and parameters (7E, 7F, 7G, 7H) describing the operating state of the dryer (15).

11. Using the model (12), (a) the control parameters (5) are determined based only on the raw material parameters (4), the state parameters (7), and the actual formulation parameters (11); or (b) the predicted actual formulation parameters (11) are determined based only on the raw material parameters (4), the state parameters (7), and the control parameters (5); 11. The method according to any one of claims 1 to 10, characterized in that properties of the intermediate product (14) that cannot be measured in-line, in particular particle-related properties of the intermediate product (14), remain unaccounted for.

12. The model (12), in particular the dynamic part (12B) of the model (12), considers the long-term effects of changes in the control parameters (5) on the actual formulation parameters (11) and / or the state parameters (7), and / or 12. The method according to claim 1, wherein the dynamic part (12B) of the model (12) is used to predict changes in state parameters (7), and based on this prediction, the basic values ​​of the control parameters (5) are adjusted, and the installation (1) is controlled using the control parameters (5) thus optimized.

13. 13. A computer program product or computer readable storage medium comprising instructions that, when said program is executed by a computer, cause said computer to carry out the method of any one of claims 1 to 12.

14. An equipment (1) for producing a formulation (2) from raw materials (3), the production including processing the raw materials (3) using a granulator (13) and drying an intermediate product (14) produced using the granulator (13) in a dryer (15), the equipment (1) comprising: sensors for detecting state parameters (7) of the equipment (1), each representing a state of the equipment (1) that influences the production; an actuator (6) of the facility (1) for acting directly or indirectly on the raw material (3); a control device (9) for controlling said actuator (6) with a control parameter (5) which is or represents a manipulated variable for controlling said actuator (6), a control device (9), in which target formulation parameters (10) representing the desired properties of the formulation (2) produced or to be produced, in particular the physical properties characterizing the particles and moisture, are preset or pre-settable for said control device (9); a model (12) based on which the control parameters (5) can be determined and / or actual formulation parameters can be predicted; Equipped with The equipment (1) comprises: the granulator (13) is coupled to the dryer (15) in such a way that the intermediate product (14) is automatically conveyed from the granulator (13) into the dryer (15) without interruption, the model (12) considers the combination of granulation using the granulator (13) and subsequent continuous drying using the dryer (15), and the installation (1) is configured such that the control device (9) determines the control parameters (5) and / or predicts actual formulation parameters (11) using the model (12) based on the state parameters (7) of the installation (1); and / or The model (12A) includes a static part (12A), and the control device (9) is configured to determine a base value of each control parameter (5) or actual formulation parameter (11) using the state parameter (7) by using the static part (12A), and the model (12) includes a dynamic part (12B), and the control device (9) is configured to optimize the base value by prediction using the dynamic part (12B).

15. 15. A system (45) comprising the installation (1) of claim 14, a device for forming the raw material (3) from a plurality of components (47), preferably powder, preferably by sieving, and / or a device for further processing (51, 55, 57) of the formulation (2), preferably tableting.

Images