Industrial extruder and process and method thereof

Abstract

An industrial smart extruder (1) and an extrusion method, the industrial smart extruder (1) serving as a conveying device that uniformly extrudes solids from a forming orifice into a viscous mass at high pressure and high temperature according to the operating principle of the Archimedes screw, the material is processed by an extruder (1) by hot or cold or warm or friction or micro-extrusion, and wherein the material extruded by the extruder (1) comprises at least a food product or a metal or polymer or ceramic or concrete or model clay. The industrial intelligent extruder (1) comprises an intelligent device with an ML or AL-based core engine that automatically controls and / or manipulates and / or optimizes the operation of the extruder (1).

Description

Technical Field

[0001] The present invention relates to the field of extruder system and process. In particular, the present invention relates to industrial extruder system, industrial extruder system comprises feeder, extruder, forming opening (die), collecting device and extruder control device, feeder feeds plastic deformation and / or viscous input material to extruder, extruder presses input material to forming opening continuously and presses out from forming opening according to the operating principle of Archimedes screw, thereby forms output material as extrudate, and the process is called extrusion. The present invention further relates to processing extruder and composite extruder. Processing extruder can be used for forming (usually piston extruder and uniaxial extruder), and composite extruder is used for chemical and physical modification (reaction, mixing, degassing etc.) of material (co-rotating type closely combines twin-shaft extruder, Buss co-kneader etc.). Extrusion process of the present invention can relate to hot extrusion, cold extrusion, warm extrusion, friction extrusion and / or micro-extrusion. Extrusion material can especially include food product, metal, polymer, ceramic, concrete or model clay. Finally, the present invention relies on technical solutions to the possibilities and challenges of applying ML-based techniques to hybrid food processing operations, in particular solutions to the possibilities of ML modeling techniques for physics-based information for food processing applications and appropriate optimization and / or automation structures. In summary, the present invention provides a valuable contribution in advancing ML-based techniques for food processing applications. Background Art

[0002] Generally, it is known for industrial extrusion that an extruder extrude an input material including a batch, applying pressure until the input material ejects from a sharpened opening (die) of the extruder. There are many different variations of extrusion equipment, but most typically, they include a hopper for holding and feeding the raw material and a hydraulically or mechanically driven device for applying pressure. Most extruders include a die for shaping the final product (extrudate) of the extrusion. The device for applying pressure typically includes a single screw or twin screw auger powered by an electric motor or a pressure head driven by hydraulic pressure. Screws commonly used in extruders include single-thread metering screws, single-barrier screws, twin-barrier screws, variable pitch threads, multi-start threads, slotted threads, and two-stage screws such as for venting extruders.

[0003] Extruders are commonly used in polymer processing, wherein extrusion is a technical process in which molten polymers are forced through a die and are used to produce parts with a fixed cross-sectional area, such as tubes and rods. Polymers are substances or materials comprising very large molecules or macromolecules composed of many repeated subunits. Due to their wide range of characteristics, both synthetic polymers and natural polymers play an important and universal role in daily life. Polymers range from familiar synthetic plastics, such as polystyrene, to natural biopolymers, such as DNA and proteins, which are the basis of biological structure and function. Both natural polymers and synthetic polymers can be produced via the polymerization of many small molecules called monomers. Therefore, their macromolecular mass produces unique physical properties relative to small molecule compounds, including toughness, high elasticity, viscoelasticity, and the tendency to form amorphous and semi-crystalline structures rather than crystals. Food polymers are polymers from edible plants, animals and microorganisms that can be used in food systems, including proteins, polysaccharides and peptides. Generally, food polymers can be classified into three groups based on their source: (1) plant-based food polymers, such as starch, dietary fiber and cereal proteins; (2) animal-based food polymers, such as dietary proteins; and (3) microbial-based food polymers, such as fungal polysaccharides. Oils and / or lipids from plants and animals can also be considered food polymers, although their molecular weight is relatively small. Therefore, food polymers represent a major area among natural polymers and play an important role in food structure, food functional properties, food processing and shelf life.

[0004] In particular, extruders are popular in food processing because they create a fast, continuous process that can be used in the food industry to manufacture a variety of food products such as snacks, breakfast cereals, granular products, pet foods, and pregelatinized powders, among others. This is a system that incorporates multiple unit operations such as mixing, kneading, cooking, forming, and cutting into a single piece of equipment. This results in a relatively simple process with high efficiency and low cost compared to other processing methods. Therefore, extrusion processing is a common processing technology with wide application in the food industry. Within the extrusion processing step, the input material constructed of food ingredients is subjected to high shear, temperature, and pressure for a short period of time. This helps to convert the ingredients from a solid powder to a molten state inside the extruder. The molten ingredients are then forced into the atmosphere through a die at the end of the extruder. The melt coming out of the extruder encounters a sudden drop in pressure, resulting in rapid expansion and a drop in temperature, thereby helping it to be converted into a cooked product. By Figure 1A schematic diagram of the transformation is given again. The final quality and texture of the extruded product depends on various factors, including the ingredient mixture and its characteristics, the extrusion process conditions and the post-processing conditions. Due to the greater flexibility of the extrusion process, it has found very diverse applications in the food industry, some of which can be seen in Table 1. Extrusion processing is largely beneficial due to its applicability in various food processes, its flexibility, its cost-reducing benefits, high productivity and quality products compared to other common food processing.

[0005]

[0006]

[0007] Table 1 shows examples of the diverse applications of extrusion processing in the food industry.

[0008] Figure 2 The diagram shows a rough block diagram of what a food extrusion processing line might encompass. The process begins with characterizing and receiving the raw materials. The raw materials used are critical to the consistency of the product at the end of the processing line. The raw materials then undergo blending and / or pre-treatment (see Fig.30 ), which can be accomplished with equipment such as ribbon blenders and pre-conditioners to ensure uniformity of the raw materials as they enter the extruder. However, mixing and pre-conditioning are optional for some products. Extrusion processing is then performed (see Fig.31 ), extrusion processing is the main cooking step in which the raw materials are converted into a cooked and shaped product. Post-extrusion processing operations such as cutting the extruded product into appropriate sizes, drying the product to the desired moisture, and seasoning or coating to provide the product with the desired flavor and taste are performed before the product is packaged. In addition to these main processing steps, there may be additional steps depending on the type of product produced and its intended use.

[0009] Extrusion processes are often defined as "hot", "warm" or "cold". Hot extrusion is performed at elevated temperatures to prevent work hardening of the material and to make it easier to push the material, especially if it is passed through a die. Warm extrusion is performed above room temperature but below the recrystallization temperature of the material. Warm extrusion is often used to achieve a balance between the desired force, ductility and final extrusion properties. Cold extrusion is performed at or near room temperature. Cold extrusion has certain advantages over hot extrusion, including no oxidation of the material, more precise tolerances and good surface finish.

[0010] As mentioned, in the food industry, extruders are widely used. Extruders are used for simple forming applications, such as pasta, or for more complex operations involving significant modification of the extruded material. Extrusion feed materials may, for example, include various solid and liquid ingredients that result in unclear conversion reactions and rheological / viscosity fields.

[0011] The main components of an extruder include a feeder, barrel, screw, and die, but many more components can be added to increase product versatility. The feeder is used to continuously feed the mixture into the extruder at a constant rate to ensure consistency. The feeder usually feeds the material by weight or by volume, and more than one feeder can be used at a time for different ingredients. The barrel encloses a screw or a set of screws. Typically, the barrel is sleeved for heating and cooling. Heating can also be achieved by providing an electric heating unit on the barrel or by steam. In a twin-screw system, the inner layer of the barrel is usually smooth, while in a single-screw system, the inner layer of the barrel can be grooved or grooved. The barrel can also have various injection ports or additional feed ports along the barrel. Injection ports can be used for water or other liquid ingredients, while additional feed ports can allow additional powder ingredients to be pressed in through one side of the barrel, so the powder ingredients experience less cooking by bypassing most of the extruder. The role of the screw is to help apply shear to the ingredient mixture and force the dough of the mixed ingredients to come out of the extruder through the die. The screw is also responsible for the pressure accumulation that occurs at the end of the extruder and the mixing of additional ingredients. The die is used to hold the material in the screw, thereby providing time for the screw to apply shear energy to the sample. The die also controls the final shape of the product and can be greatly changed. The last necessary component of the extruder is the motor, which provides the energy required to rotate the screw. Other optional components may, for example, include a pre-processor, which can be used to pre-hydrate the ingredient mixture before feeding it into the extruder and pre-cook the material in some cases. A die cutter is also often used to help cut the final extrudate coming out of the die.

[0012] There are two main types of extruders: single screw and twin screw (co-rotating and counter-rotating). These extruders are available in a wide range of screw diameters and length / diameter (LID) ratios with or without steam pretreatment. It may not always be easy to optimally select and design an extruder for a specific application given current process technology limitations. The various known extrusion processes can be differentiated based on their functionality and operating parameters by addressing different process requirements and / or by their logical approach to selecting or designing the process. It is fundamental to understand the interaction between material properties, rheology, conversion reactions, and system design and operating parameters.

[0013] For example, a prior art discussion may begin with the use of single screw extruders for plasticizing / melting synthetic polymers, since most extrusion theory development is based on this application. Comparing plastics to food extrusion can then be used to learn from and properly analyze the single screw food extrusion process. Co-rotating and counter-rotating twin screw extruders are also found in various forms and can be compared to single screw extruders and to each other.

[0014] Food extruders utilize thermal and mechanical energy. Understanding energy consumption and input requirements is very important for improved performance and economical system design. Water is a common ingredient in almost all food extrusion formulations. A portion of the water can sometimes be applied in the form of steam to supplement the total extruder energy. This option may have a significant impact on extruder selection, design and performance, as well as on product characteristics. Combined extrusion cooking and shaping of unexpanded products is a common process for making snacks, cereals and other products. Accelerated cooling before the final shaping step is critical in this operation. This is usually achieved by evaporating a portion of the liquid water in the extruder or in the exhaust stage between the cooking utensil and the shaping extruder. When extruded products are developed on a smaller production basis, expansion becomes part of the entire extruder selection and design process. Understanding the expansion process is very important for identifying its important factors, limitations and supplementary process options. Finally, it should be noted that the design parameters and operating variables of the extruder are usually interactive, affecting extruder performance and product characteristics. Therefore, people can choose more than one system design and operation to make the same product. The best choice may be determined by availability, flexibility and economy.

[0015] Modeling plastic and food extrusion operations

[0016] Non-food polymers can be divided into three main groups: thermoplastics, thermosetting plastics and elastomers. Thermoplastic materials generally solidify when they are heated and when they are cooled. The chemical composition of thermoplastic materials generally does not change during the extrusion process. Thermosetting plastics undergo a cross-linking reaction when the temperature is increased beyond a certain limit. This leads to the formation of a three-dimensional network, which remains intact when the temperature is reduced. The cross-linking reaction is irreversible, and therefore thermosetting plastics cannot be recycled as thermoplastic materials. Elastomers or rubbers are materials that can undergo very large deformations when subjected to shear forces but return to their original shape when the applied force is removed.

[0017] In the prior art, the development of polymer extrusion technology has been mainly focused on the use of extruders for thermoplastics. In most cases, only a single polymer component is extruded. The process involves conveying, plasticizing (melting) and metering the melt to the die.

[0018] Two types of polymerization reactions, namely "addition" and "condensation", produce plastic polymers. In the first type, single monomers, such as ethylene, directly add to each other without changing the composition and form long chain molecules. In the second type, two or more monomers react with each other to form linear chain molecules with the elimination of a small part of them (usually water). The reactive monomers in this reaction type usually carry two functional groups that can react with each other continuously. Examples of polymers of this type are polyamides (nylons) and polyesters. All polymers in linear or slightly branched polymers produced by addition or condensation reactions are thermoplastic. This means that they can be softened repeatedly at elevated temperatures and solidified again by cooling. No chemical changes occur during these thermal treatment processes.

[0019] Extruders are commonly used to convey solid polymers, melt, mix and form finished product shapes such as sheets, tubes and rods. Extruders are also used to compound and produce plastic raw materials, such as pellets for other processing operations. In addition to continuous operation, extruders are also used in a semi-continuous manner in the case of injection molding machines by reciprocating the screw forward at the beginning of the mold filling cycle.

[0020] Extruded food ingredients are mainly composed of biopolymers of carbohydrates (starch, fiber sugars, etc.) and proteins. Other ingredients such as water, oil / fats, minerals and vitamins are also included in the feed mixture. The structure of biopolymers is not well defined and varies with the source. In food extrusion, the process usually involves feeding a mixture of solid and liquid ingredients. These ingredients react in many cases in a complex and unclear process involving irreversible changes in physical and chemical structures. Inside the extruder, the ingredients form a fluid material due to any of the following: (1) Low temperature dough formation process in the presence of sufficient fluid (water and / or oil). Water is always present and acts as a plasticizer, thereby reducing the glass transition and melting temperature. If sufficient energy is provided, the components of the formed dough can undergo structural modifications. For example, if a certain temperature range is exceeded, starch granules will undergo a gelatinization / melting process. The latter may lead to a significant viscosity increase due to the increase in the water absorption capacity of the gelatinized starch; (2) At low water / oil concentrations, the softening and melting processes of biopolymers (carbohydrates and proteins) can be achieved by applying sufficient energy. This can be in the form of solid friction and mechanical energy, viscous heat dissipation of steam energy and / or heat transfer through the barrel. Typically, the food material also undergoes irreversible physical and chemical changes in this case.

[0021] The heated melt behaves as a non-Newtonian pseudoplastic material in most food extrusion situations. Its viscosity decreases with increasing shear rate and temperature. Its viscosity is also reduced by increasing the concentration of water and other viscosity reducing agents.

[0022] Food extrusion can be divided into two general categories: forming and cooking. For forming applications, low shear extruders are basically used to mix and form the desired product shape with minimal energy input. Product applications include pasta, cold formed snacks and other unexpanded pre-cooked pellets. The screw usually has a deep channel (0.2D to 0.3D) without a compression ratio and is operated at a relatively low speed (less than 50rpm).

[0023] Cooking extrusion applications typically utilize medium shear single screw co-rotating extruders and high shear twin screw co-rotating extruders. A large amount of energy can be provided to the product through viscous heat dissipation, heat transfer through the barrel, and sometimes through steam injection. For single screw extruders, the screw runs at a relatively high speed (>100 rpm), with a shallow channel depth and a compression ratio of up to 4:1. These extruders are used for cooking as well as forming expanded products.

[0024] Both single screw extruders and twin screw extruders are used for plastic and food extrusion. Twin screw extruders have significant advantages over single screw extruders in feeding, mixing, heat transfer, residence time distribution and pumping performance. However, twin screw extruders are more expensive. The main areas of twin screw application in polymer extrusion are: (i) extrusion of heat sensitive materials, such as profiles of polyvinyl chloride (PVC); and (ii) complex and difficult polymer processing operations that require good mixing, devolatilization and chemical reactions of various compounds.

[0025] Single screw extrusion

[0026] Figure 4 The a / b shows the following examples: (a) thermoplastic single screw extruder, (b) thermoplastic single screw. Figure 5 A description of screw terminology is provided using single screw extruder screw elements.

[0027] Twin screw extrusion

[0028] With respect to twin screw extrusion, the prior art discussion will be limited to intermeshing twin screw extruders. There are two types based on the way one screw rotates relative to the other screw: co-rotating extruders and counter-rotating extruders. One major difference between single screw extruders and twin screw extruders is the type of transport that occurs within the extruder. Single screw extruders are completely dependent on frictional drag flow in the solid conveying zone and viscous drag flow in the melt conveying zone. Transport in a twin screw extruder is less dependent on the frictional characteristics of the material due to the action of the second screw in the intermeshing zone. This provides a degree of positive displacement transport. Intermeshing counter-rotating twin screw extruders provide the greatest positive displacement action.

[0029] The channels in a co-rotating twin-screw extruder cannot be physically completely closed in the intermeshing area (see Figures 6 to 8 ). This is due to the nature of the screw rotation. Due to the wiping of the screw surfaces in the intermeshing area, the material is wiped off from one screw surface, changes direction and moves to the other screw (see Fig. 9 ). The pressure difference within the extruder acts on the material and can cause a backward movement since the channel is open in the axial direction.

[0030] In a fully intermeshing counter-rotating extruder, the screw channels are closed in both the radial and axial directions. The material is physically prevented from moving to the other screw by the threads of the second screw in the intermeshing region. Thus, for each screw rotation, it is forced to move one screw pitch (along each screw) as a "C" shaped closed chamber (see Figure 6 and Fig. 9 ). As a result, these extruders are less susceptible to pressure flow and are able to produce higher die pressures than other extruders under similar conditions. Material mixing is limited by the flow distribution within the closed chamber of the "C" shape and the leakage flow between the screw and barrel components.

[0031] The material flow in a co-rotating twin-screw extruder follows a figure "8" distribution with corresponding changes in direction and surface renewal. This results in a relatively uniform shear stress distribution around the screw (see Fig.10 ). The screw configuration typically includes mixing and flow restriction elements. This creates a complex flow pattern that results in good mixing and heat transfer, large melting capacity, and good melt temperature control. Flow pattern complexity makes it more difficult to model and scale up. In the prior art, successful scale-up has been partially achieved through modular extruder designs, where the screw configuration can be easily adjusted to obtain the desired performance characteristics.

[0032] The shear stress distribution around the screw in a counter-rotating extruder is not uniform (see Fig.10 ). Material caught in the intermeshing area pushes the two screws outward toward the barrel, which can cause excessive wear. This limits their operation to melt extrusion at relatively low viscosities and at low screw speeds. Their strength is consistent with their highly positive displacement delivery. Twin screw extruders are typically starved fed, and the feed system controls the output of the twin screw extruder. The degree of screw filling will depend on the flow rate, material viscosity, screw design, screw speed, and the pressure distribution created by the restricting screw elements or die assemblies. Therefore, the mechanical energy input depends on the extruder design, shear rate, and material viscosity. It should be noted that the material flow distribution within a co-rotating twin screw extruder is complex and technically more difficult to capture by modeling or simulation than that of a single screw extruder.

[0033] In the prior art, twin screw extruders were first used for food cooking applications in the 1970s. The use of twin screw extruders has increased at the expense of single screw extruders and due to the widespread expansion of food extrusion applications. As previously mentioned, co-rotating twin screw extruders have greater positive displacement and mixing capabilities than single screw extruders. This has enabled the development of improved processing capabilities and flexibility and more consistent food product quality. Twin screw extruders have been adopted in low viscosity and medium viscosity and high viscosity applications. Modular extruder design and good process control have broadened its use to applications requiring multiple processes and functional areas. These applications include multiple liquid injections, downstream solid feeding and exhaust ports. Twin screw extruders can be designed with relatively long barrels, thereby providing longer residence times required for certain conversions and reactions. The flexibility in extruder design has also accelerated the development and optimization of new and more difficult products, despite the fact that their exact flow behavior is not well understood. Fig.11 An example of the possible capabilities of a twin screw extruder for a multi-process extrusion system is shown. It shows two dry mix feeders, liquid injection ports and direct steam injection ports, side feeders for other solid components, exhaust port with exhaust plug, injection port for heat sensitive liquid ingredients, start / throttle valve, die plate, and die face cutter assembly. In reality, most extrusion systems are much simpler and may not include Fig.11 Some of the units shown in .

[0034] Counter-rotating extruders have been used for low viscosity applications such as candy and spirits that require high pressure molding capabilities. They typically rely heavily on barrel heat transfer to heat and cook the material. The material captured in the C-shaped chamber can be heated to high temperatures without steam blowing back to the feed port. The pressure within these chambers can be raised and maintained to a level higher than the corresponding local steam saturation pressure.

[0035] Considerable work has been performed in the prior art to provide modeling and simulation structures that capture co-rotating twin screw extruders. Most of these efforts focus on establishing empirical relationships associated with the effects of feed components and operating conditions on product properties. Twin screw designs typically incorporate mixing disks, paddles, or reverse screw elements. The additional complexity of food components makes the task of detailed engineering modeling even more difficult. In the prior art, only very few modeling approaches exist. For example, there is prior art simulation modeling of twin screw co-rotating extruders, however, the modeling structure is only unidirectional and only considers non-Newtonian and non-isothermal melt rheology.

[0036] forming

[0037] The molding process is understood as the process of passing the melt through a specially designed die opening. A certain pressure drop occurs on the die depending on the die geometry, melt viscosity and flow rate, where ΔP≈Qμ / K, where K is the die conductance factor equal to the inverse of the die resistance. However, this relationship generally assumes that the laminar flow through the die does not have material slip on the metal surface. Slippage leads to a pattern similar to plugged flow and can be promoted by using smooth and low-friction die materials. Die heating may also reduce the friction factor, thereby further increasing slippage. Material slippage is expected to result in a lower pressure drop on the die. This may affect the total extruder filling, mixing intensity and energy input. From the perspective of product molding, the main advantage of die slippage is the formation of a smooth and complete product surface. This is why Teflon inserts are often used in pasta product molds.

[0038] Rheological analysis of food melts at different shear rates (volume output) is more complex than that of thermoplastics. One needs to consider the changes in melt properties with changes in feed rate because it affects the time-temperature history in the extruder.

[0039] In most food extrusion operations, the extrudate expands after leaving the die, e.g. Fig.12 The large and complex molecular structure of the polymer creates elastic properties. In the prior art, this is considered to be the main reason for this phenomenon. Swelling is a form of elastic recovery from the deformation of the polymer after exposure to the mold. Swelling increases with the sudden mold flow entering and through the shortening of the mold platform length. Swelling also increases with increasing shear rate within the mold. In the prior art, in a circular mold, the shear rate is usually modeled by approximation as γ = Q / πr 3 .

[0040] In the case of complex extruded food product shapes, the design and development of prior art dies has been based more on a logical and practical approach than on an analytical approach. These shapes have been further developed in cold extrusions, such as pasta and semi-finished products. Streamlining of material flow in cooked extrusion applications is important to avoid material stagnation. Such materials can lead to quality defects, process instability and potential blockages in small die areas.

[0041] mix

[0042] The basic mechanism of mixing in a polymer extruder is convective motion in the laminar flow region. Mixing action usually occurs through shear and elongational flow. It is called distributive mixing when the mixed components do not show a yield point. Distributive mixing can be described by the degree of deformation or strain to which the fluid elements are exposed. On the other hand, dispersive mixing is similar to milling action and addresses the decomposition of solid particles and agglomerates. Actual stress is important in this form of mixing to overcome the yield stress of the solid components. Distributive mixing is always present when dispersive mixing occurs, but dispersive mixing is not necessarily always present when distributive mixing occurs.

[0043] Distributive mixing in an extruder is usually simulated by determining the velocity profile occurring in the screw channel. In most simulation methods, the fluid is considered to be Newtonian, the components have the same flow characteristics, and the flow through the flight gap is neglected. Mixing performance improves with increasing pressure to resistance flow ratio (throttling ratio). In a choked feed extruder, this will reduce the extruder output. In all cases, this will increase the mean residence time and broaden the residence time distribution (more back mixing).

[0044] Special mixing elements have been designed for single screw extruders to promote distributive mixing. Types of distributive mixing elements include pins, dulmage, saxton, pineapple, slotted threads, and cavity transfer mixing sections. Quantitative analysis of these mixing elements is quite difficult, and the development of these devices has been mostly empirical using experimental evaluation methods. When used in thermoplastic single screw extruders, they are usually placed at the discharge end of the metering section. Modular single and twin screw extruders use kneading paddles, cutting thread screws, and discs to promote distributive mixing. These elements are placed in any logical position in the extruder.

[0045] It is known that dispersive mixing is critical in compounding and some food extrusion applications. The decomposition stress of the agglomerates depends on the size, shape and properties of the agglomerates. The stress acting on the agglomerates will depend on the flow and rheological properties of the polymer field. The higher the viscosity, the stronger the dispersive mixing will be. High-speed solid-solid mixing may also cause a higher level of agglomerate decomposition compared to liquid-solid mixing. Several dispersive screw elements have been used in extruders. These include Union Carbide, Egan, Dray and Blister rings. In all these cases, the agglomerates must pass through the specially designed gap between the mixing element and the barrel. The agglomerates are subjected to high shear stresses, which decompose the agglomerates and enhance mixing.

[0046] Twin screw extruders can be provided with wide kneading paddles to promote dispersive mixing. It is not possible to create a tight gap dam area in an intermeshing twin screw extruder with side-by-side equal diameter disks. In this case, the maximum disk diameter is equal to the center distance between the two screw shafts. The resulting tolerance / clearance with the barrel will be equal to half the screw channel depth. A tight gap dam can best be created by placing two sets of intermeshing disks (following each other) on the screw shaft (see Fig.14 ).

[0047] Modeling the pressure development in an extruder

[0048] As mentioned previously, many of the food single-screw and twin-screw extruders are typically starved fed. Their output is determined by the feed system. Therefore, the screw in the solid conveying section partially keeps filling the negligible mechanical work done on the feed material. Screw filling occurs in successive sections due to the following reasons: (i) the conveying capacity of the advancing screw or kneading element is reduced, or (ii) obstacles are placed in the form of kneading elements, disks, reverse screws or die plates. As the flow resistance increases, the feed materials (solids and liquids) are compacted, mixed and converted into fluids / melts through a hydration process accompanied by solid-solid / solid-fluid friction and dissipation of mechanical energy. This conversion / melting process may involve significant temperature increases depending on the friction field generated and the viscosity of the material.

[0049] The length of the conveying screw section to generate the necessary pressure to overcome the restriction is affected by the screw design, effective material viscosity and screw speed. The pressure gradient for a Newtonian fluid can be modeled using the following relationship: Among them, K 1 is a constant and is affected by the number of screw flights, the degree of intermeshing, and the clearances between the screws and between the screw tip and the barrel; D is the screw diameter; N is the screw speed; μ is the material viscosity; H is the flight height; and is the screw flight helix angle. From the last relationship it can be seen that for a given extruder, the length of the melt pumping section length decreases with increasing screw speed, melt viscosity and decreasing screw channel depth and helix angle. This equation applies to the case where the material is completely or partially in fluid (melt) form (serving as the continuous phase).

[0050] For Newtonian fluids, the pressure gradient across the resistance segment made of kneading elements, disks, reversing screws or die plates can again be modeled by the relationship ΔP≈Qμ / K, where Q is the melt volume flow rate and K is the conductance factor of the restriction segment. For the reversing kneading / screw element segment, K will be affected by the formed helix angle, channel depth, screw gap and screw diameter. In steady state, this segment will always be full and forward flow of material occurs by the pressure drop in this segment. The pressure flow should be larger than the resistance flow component of the restriction element of this segment.

[0051] For a discharge die obstruction such as a circular orifice, K can be modeled, for example, as K = π·r 4 / 8L, where r is the mold radius and L is the mold platform length. The pressure drop within the resistance section increases with increasing flow rate, viscosity, and restriction of the section. Fig.15 An exemplary twin-screw extruder is shown, showing simulated pressures and temperatures within the extruder.

[0052] Modeling of energy consumption and supply in extruders

[0053] Modeling energy consumption and supply throughout the various processes and sources within an extruder is complex. The following points must be considered:

[0054] (a) Energy supply in the extruder: The energy supplied in the extruder is mainly consumed by the product. Some unavoidable losses also occur in the barrel, gearbox and motor. The energy consumed by the product is usually modeled by the following changes: (1) Enthalpy rise q h (e.g., in kW). In cooking extrusion applications, enthalpy increase accounts for the largest amount of energy consumed; (2) Reaction / conversion heat such as heat of starch gelatinization and molecular decomposition. Gelatinization energies have been determined for different starches and are generally estimated to be in the range of 10 kJ / kg to 20 kJ / kg. These values ​​are generally measured at excess water concentrations, taking into account only the gelatinization endotherm. In high viscosity extrusion applications (at low moisture concentrations), starch and other biopolymers undergo a fractionation (depolymerization) process that is also expected to consume some energy. This energy is difficult to measure and difficult to quantify; (3) Potential energy q P (eg, in kW).

[0055] The material kinetic energy is usually assumed to be too small and is therefore neglected. Energy losses may be contributed by: (1) Barrel energy losses due to circulating liquid in the barrel jacket q cl(e.g., in kW); (2) barrel energy losses in the form of convection cooling through the surrounding air; (3) energy losses due to the gearbox and motor. This comes in the form of forced cooling to remove the heat generated and keep the gearbox and motor at acceptable temperatures. The losses can be estimated by calculating the applied cooling load. The losses are typically in the range of 5% to 15% of the motor energy, depending on the size and operation of the extruder.

[0056] (b) Energy supply: Energy is supplied to the material from the extruder motor, barrel heat transfer and in some cases by direct steam injection.

[0057] (c) Mechanical energy input: Mechanical energy is provided by the extruder motor. Mechanical energy is consumed primarily by the material in the filling section of the extruder through friction and viscous heat dissipation. For Newtonian fluids, the mechanical energy input (E) can be expressed by ∫dE≈∫c μ ·N 2 dx, where C is a constant representing the screw design. Food materials typically exhibit pseudoplastic non-Newtonian flow behavior. Together with the effects of moisture (M) and fat (F) content and temperature (T), viscosity is often empirically calculated using μ 0 The reference (initial) viscosity and γ are modeled as the shear rate within the screw. The latter is linearly proportional to the screw speed.

[0058] It can be observed that the viscous heat dissipation can be increased by increasing the length of the screw filling section, the material viscosity and the screw speed. It should be noted that increasing the screw speed leads to a decrease in the length of the melt pumping section. On the other hand, the resistance sections (discs, reverse kneading elements and screws) will always be full and their filling length is constant regardless of the screw speed or feed rate.

[0059] In the prior art, the total specific mechanical energy input (kWh / kg) for a direct current (DC) motor with a constant maximum torque is estimated or approximated using a modeling structure, for example, as S=(%N·%T·p) / F, where N is the maximum extruder screw speed (rpm), T is the maximum motor torque (Nm), P is the total installed motor power (kW), and F is the feed rate (kg / h). The screw speed and torque are proportional to the armature voltage and current of the motor, respectively. For an alternating current (AC) motor, the actual power should be obtained from a wattmeter.

[0060] (d) Heat transfer through the barrel: The heat transfer rate (kW) to or through the extruder barrel can be modeled using the modeling structure a h =U·A·ΔT lm Model or approximate, where U is the convective heat transfer coefficient (kW / (m2℃), A is the internal barrel surface area (m 2 ), and ΔTlm It is the logarithmic temperature difference (°C) between the internal barrel surface and the material in contact.

[0061] The heat transfer coefficient in the non-filled section is relatively small because there is no intimate contact with the surface. The heat transfer coefficient is affected by the physical properties of the material and the mixing pattern. The heat transfer coefficient generally increases with decreasing viscosity and increasing screw speed. Levine and Miller (2007) provide a review of published research work on barrel heat transfer. The heat transfer coefficient has been estimated to be around 50 W / m 2 ℃ to 300W / m 2 ℃ range.

[0062] The role of barrel heating becomes more important and significant in low viscosity extrusion cooking applications, such as those containing high moisture, sugar and / or fat concentrations. The rate of mechanical energy input is reduced in such applications. In high melt viscosity applications, motor mechanical energy is usually the primary source.

[0063] Heat can also be transferred from the material by applying cooling to the barrel jacket. Product cooling causes the melt viscosity to rise, resulting in an increase in the rate of mechanical energy input. Therefore, the net product cooling is lower compared to the total cooling load taken out of the barrel.

[0064] The barrel heat transfer per unit output may be greater in a small extruder than in a larger extruder. In both cases, barrel temperature control is important even when the heat transfer rate is not significant. As mentioned previously, the coefficient of friction of a material with a metal surface may be affected by the surface temperature.

[0065] (e) Steam injection into the barrel: Water is usually a natural component of the total food formulation. A portion of the total amount of water can be injected directly into the extruder in the form of steam. This can bring a large amount of energy, thereby supplementing and / or replacing a portion of the motor and barrel heating energy. In the prior art, for example, the steam energy q s (kW) modeled using the relation q s =m s ·λ s q s The steam mass flow rate is usually injected in the non-filled section of the extruder. Control of the steam flow rate is important for achieving steady-state operation. It is also necessary to provide a highly filled section between the injection port and the dry mix feed port to avoid steam blowback. This can be achieved by using screw elements that reduce the conveying capacity (small pitch or cutting flight screws) or restricting disk "shear locks".

[0066] Modeling the Effects of Screw / Barrel Wear on Extruder Performance

[0067] Extruders act as pumps and thermomechanical reactors. These functions are affected by the extruder's screw design, including the various gaps within the extruder. Under use, the screw elements and barrel materials undergo a wear process. The wear rate will depend on the abrasive, adhesive, and corrosive properties; the construction materials of the components; and the operating conditions. An increase in the gap size of the forward conveying section reduces the extruder's pumping efficiency and increases the degree of back mixing and residence time. In this case, the total mechanical energy input to the product may increase. If the extruder is operated in a choked feed mode or a near-choked feed mode, this may also reduce the extruder output.

[0068] In addition, restricted sections within the extruder and die assembly may also experience wear. The resulting larger clearances will in this case reduce the flow resistance of these sections, thereby requiring a reduction in the pressure drop across these sections. This may reduce the extruder fill length of the traveling pumping section and the overall energy input to the product.

[0069] For heat sensitive materials such as food products, larger gaps and possible material stagnation can potentially lead to product degradation and quality defects, possible die clogging, and process instability.

[0070] Adjustment of product formulation (particularly the rate at which water is added) and operating conditions can in some cases compensate for gradual wear and extend the service life of extruder components. The screw profile can also be changed to adjust the melt filling section length. In the case where the extruder length exceeds the necessary, one can move the filling (working) section up or down along the screw to use the unworn section of the barrel liner. However, this requires sufficient processing knowledge about the effect of the screw configuration on process and product performance. Eventually, a situation will be reached where product and / or output specifications cannot be met, and component replacement will have to be performed. In the case of new components, if the extruder conditions have changed during the component wear life, the extruder conditions may have to be adjusted again.

[0071] Exhaust in the extruder

[0072] Venting within a cooking extruder can be very useful for many purposes, such as accelerating cooling, removing certain volatile components, and affecting certain changes in the product. A specially designed screw is required to allow venting in the non-filling section. Typically, a high volume flow screw section is positioned at the vent port after a restrictive / low volume flow screw section that acts as a gas tight seal (see Fig.16 ).

[0073] Melt rheology and temperature affect the behavior of the melt and the extent to which volatiles are removed from the venting section. Relatively low viscosity melts (e.g., those with high initial moisture content) can be successfully vented. In cases where the initial moisture / fat content is reduced and / or vapor release is excessive, the vented melt may produce a highly expanded mass that is difficult to convey forward. Using a plugger (see Fig.17 ) may help in some cases.

[0074] When such a cooked semi-product is applied, considerable product cooling is required to avoid product expansion and to achieve a good die face cut. The optimal extrusion process design will depend on the nature of the formulation and the desired degree of cooking. If this can be achieved at relatively low temperatures, cooking, cooling and forming can be performed in a single extruder with or without the use of venting ports.

[0075] In applications with relatively high temperatures (e.g. above 135°C in the cooking section) and at high production capacities, it is often preferred to use a "two extruder" system. The first extruder cooks the material. Exhaust and cooling are achieved between the two extruders. The second extruder compresses the expanded material and forms the cooled product into the desired shape. Thus, the two extruders can be designed and operated independently. In this way, the cooking extruder and the molding extruder are utilized to their maximum output and efficiency without having to compromise as is the case with a single cooking / molding extrusion system. The molding extruder is typically a single screw type similar to those used in pasta product applications.

[0076] Conclusions on the simulation and modeling process for extruders and their usability in extrusion automation and robotic operation of industrial extruders

[0077] Extruders are common devices in the plastics, metal and food processing industries, and the use of the extrusion process is particularly widespread in product manufacturing and food processing. In particular, extruders are popular in food processing because they create a rapid, continuous process that can be used in the food industry to make a variety of food products such as snacks, breakfast cereals, pellet products, pet foods, and pregelatinized powders, etc. This is a system that incorporates multiple unit operations such as mixing, kneading, cooking, forming, and cutting into a single piece of equipment.

[0078] Generally, the main extruders can be classified as single screw type or twin screw type, the single screw type is widely used for general polymer processing, and the twin screw type is used for compounding various fibers, fillers and polymer blends before final molding. Single screw extruders include only one screw housed in a barrel, which is usually of a grooved or grooved design. In addition, the screw in a single screw extruder is usually designed with a reduced pitch to produce compression. The amount of reduction in pitch is called the compression ratio.

[0079] Twin screw extruders can be further divided into two types based on the interaction of the two screws: intermeshing twin screw extruders and non-intermeshing twin screw extruders. In the twin screw extruder series, it has been found that fully intermeshing counter-rotating twin screw extruders have the best pumping capabilities due to their positive displacement characteristics. Therefore, a set of screws in a twin screw extruder can rotate in the same direction or in the opposite direction. Co-rotation is more frequently used because it can apply more mechanical energy to the material compared to counter-rotating screws. Twin screw extruders are more commonly used in the food industry due to their wide range of operating conditions and their ability to manufacture a wide range of food products, even though twin screw extruders require higher maintenance costs compared to single screw extruders. Twin screw extruders have a wide range of flexibility for handling diverse ingredients and have higher productivity compared to single screw extruders. Twin screw extruders can operate with a wider range of moisture contents, which is a disadvantage of single screw extruders. Pretreatment systems can be used to extend the capabilities of both single screw extruders and twin screw extruders. Twin-screw extruders offer higher mixing efficiency, self-wiping capabilities to prevent residue buildup, and relatively faster and more uniform heat transfer from the barrel to the ingredients.

[0080] In principle, extrusion is a multiple-input and multiple-output (MIMO) system. Figure 3 The inputs and outputs associated with this concept are shown. The various extrusion processing parameters can be broadly classified into three categories: (1) independent parameters (input parameters), (2) system parameters (dependent parameters) and (3) product characteristics (output parameters). Independent parameters are parameters that the extruder operator can directly control. These parameters include raw material (ingredient) characteristics and extrusion operating parameters such as feed rate, barrel temperature, screw configuration, screw speed, die size, etc. By modifying the independent parameters, the operator can achieve changes in system parameters and final product characteristics. System parameters are parameters that the extruder operator cannot directly control, but the operator can influence the system parameters by changing the independent parameters. Average residence time, residence time distribution, back pressure, motor torque and specific mechanical energy applied to the material are measurable system parameters. These parameters affect the final product characteristics, but can only be modified indirectly by changing the independent parameters. Product characteristics are parameters that help describe the quality of the final extruded product. This can include physical properties (expansion ratio, density, etc.), chemical properties (water absorption and solubility, etc.) and sensory properties (crispness, crispness, texture, etc.). In order to manipulate, optimize and / or control an extrusion process, for example to consistently produce a food product and / or optimize energy consumption, there is a need to capture, parameterize and control a MIMO system.

[0081] In the modeling of food and polymer extrusion processes, the models are usually distributed parameters or local lumped parameters as described above, based on deterministic approximations of transport phenomena. For engineering, lumped parameter methods are usually used. In the prior art, the main goal of engineering design of extrusion is to predict the pressure along the machine and the average process material temperature according to the operating conditions for a given screw / die geometry. In these model structures, the screw channel is divided into short segments, where the input data comes from the calculations in the previous segment and the output data from the current segment is the input data for the next segment. In the segment, the local parameters are assumed to be constant. This local lumped parameter concept is particularly used when dealing with plasticization processes, such as extrusion, where, in addition to the melt flow, solid transport and melting are also modeled.

[0082] Computer modeling is often used to answer technical questions about food and polymer extrusion. Computer modeling is used to study the effects of changes in input parameters on material / process output and to show trends in the response of materials to hundreds of potential input data configurations. Computer modeling is used to obtain a process diagram for identifying the required process parameters based on the target material and the selected screw profile, i.e., the distribution of thermomechanical results (temperature, pressure, viscosity, filling) along the screw, detailed energy balance of the process, residence time analysis, etc. As another technical application, the expansion module of the software is used to transfer the process to a machine with a larger or smaller diameter (scaling of the process). Starting from the model process, the geometry of the target design and the associated process parameters are technically simulated / modeled based on the transfer rules. However, there are many shortcomings in actual use. An important point is that the global modeling structure that describes the complete extrusion process including solid transport, polymer melting and melt flow is based on a relatively simple 1D or 2D model and does not allow for the study or technical capture of more details of the process. In addition, the physics of extrusion is not always fully understood and / or is often too complex to be captured and / or simulated by an achievable modeling structure. However, even if the physical properties would be known and could be modeled, modeling approaches often fail technically due to the complexity and uncertainty of multi-parameter processes, thus achieving only poor accuracy, or requiring huge processing capacity to allow operation within an acceptable time frame. Summary of the invention

[0083] The object of the present invention is to provide an industrial system and method for the automation and / or optimization control of industrial extrusion processes. The system should be able to optimize food and / or polymer processing by extrusion techniques for all of the numerous unit operations involved, including mixing, kneading, shearing, cooking, forming and shaping. It should be noted that the extrusion cooking process alone is usually a complex, high-temperature, short-time process in which the food material is cooked in a tube by a combination of moisture, pressure, temperature and mechanical shear, resulting in molecular transformation, gelation, protein denaturation and bond breaking, thereby producing products with new shapes and textures. The system should be able to further optimize the extrusion process, resulting in lower energy consumption, reduced anti-nutritional factors, improved product microbiological safety and thus increased consumer acceptance. In addition, the system should also be able to automatically adjust and optimize the raw material properties including thermophysical properties and their flow properties under set conditions in the extruder, which is crucial for controlling the extruder behavior to obtain high-quality final products. The system should also be able to take into account the thermophysical properties of the material, such as density, specific heat, thermal diffusivity and thermal conductivity, which are usually the basis for the design and optimization modeling of food processing operations involving heat and mass transfer. The system should also be able to process input measurement parameters that are physicochemical properties that are often important for capturing and characterizing or measuring changes in processing materials, such as food materials, during extrusion, and can give explanations that account for thermal and mechanical effects and their changes. Finally, the system of the present invention should be able to capture the effects of extrusion variables (feed moisture content, screw speed and barrel temperature) on the approximate, thermophysical, physicochemical, extrudate, color, texture and sensory properties of processing materials, such as food and / or polymeric materials, and react dynamically to them.

[0084] According to the invention, these objects are achieved in particular by the features of the independent claims.In addition, further advantageous embodiments can be derived from the dependent claims and the associated description.

[0085] A generally known extruder system comprises a feeder, an extruder, a forming opening (die), a collecting device and an extruder control device, wherein the feeder feeds a plastically deformable and / or viscous input material to the extruder, the extruder continuously presses the input material to the forming opening and presses it out from the forming opening to form an output material as an extrudate, wherein the collecting device collects the extrudate for further processing, wherein the extrusion process is manipulated in a controllable manner by the extruder control device during operation, and wherein the extruder control device comprises a programmable logic for setting and adjusting operating setting parameter values ​​208 of the operating unit 105 of the feeder 101, the extruder 102, the forming opening (die) and the collecting device.

[0086] Such known extruder systems are manually controlled or semi-automatically controlled, in particular the process parameters for controlling the system are set with the aid of an extruder control device based on the experience and expertise of the operator of the extruder system. Due to the increasing demands on the material characteristic parameters / material properties of the extrudate on the one hand and energy-saving operation of the extruder system on the other hand, such extruder controls which are manually or semi-automatically operated by an operator do not ensure an optimal control of the extrusion process and the material characteristic parameters. For example, plant-based protein extrudates have very high demands and tight tolerances on the material characteristic parameters, since these products are directly compared to meat / fish products, and on the other hand the operating control parameters have a non-linear relationship with the input parameters for controlling the extrusion process. Furthermore, the presence of knowledge and experience of the extruder system requires the presence of an operator, in particular a change of the operator leads to at least a partial loss of the operator's knowledge and / or experience.

[0087] According to the invention, this object is achieved in particular by the elements of the independent claims. Further advantageous embodiments are evident from the dependent claims and from the description.

[0088] In particular, these objects are achieved by an extruder system in the present invention, the extruder system comprising a feeder, an extruder, a forming opening (die), a collecting device and an extruder control device, wherein the feeder feeds a plastically deformable and / or viscous input material to the extruder, the extruder continuously presses the input material to the forming opening and presses it out of the forming opening to form an output material as an extrudate, wherein the collecting device collects the extrudate for further processing, wherein the extrusion process is manipulated in a controllable manner by the extruder control device during operation, and wherein the extruder control device comprises a programmable logic for setting and adjusting operating setting parameter values ​​of an operating unit 105 of the feeder, the extruder, the forming opening (die) and the collecting device, wherein the extruder control device further comprises a digital controller for sending signals to the programmable logic and manipulating the programmable logic, wherein, in order to manipulate the programmable logic, the digital controller captures input parameter values ​​including at least process parameter values ​​and / or operating setting parameter values ​​and / or material characteristic parameter values ​​and / or environmental measurement parameter values, wherein, The input parameter values ​​include sensory parameter values ​​measured by sensors associated with the feeder and / or the extruder and / or the forming opening and / or the collecting device, because the extruder control device includes a repository storage unit (with an adaptive digital database 504), which stores a plurality of selectable structured data records for storing digital recipes, each of the selectable data records including at least material property parameters of the input material and target material property parameters of the extrudate 300 and / or initial operating setting parameters, the initial operating setting parameters giving the initial settings of the operating setting parameters 201 for the operation 105 of the extruder system 1, wherein the input parameter values ​​200 also include parameter values ​​of the selected data record 507, and the extruder system 1 includes digital signals for manipulating the programmable logic 501 and the associated operating unit 507 by means of the digital controller 502 so as to controllably and manipulably extrude the extrudate 300 having a material property parameter value 202 within a predetermined tolerance range of a predetermined target parameter value 206.

[0089] The digital controller has a recipe database for extrusion data with predetermined material properties, which ensures the generation of extrusion data predetermined in the recipe for setting and adjusting the values ​​of the operating setting parameters of the extruder control device. Preferably, the recipe includes initial operating control parameters for starting the extruder system so as to start with optimal process parameters and thus ensure a stable and energy-saving extrusion process. In addition, the input parameters include environmental measurement parameters for characterizing the environment of the extruder system, such as temperature, humidity and weather forecast data of the environment of the extruder system. The environmental measurement parameters are taken into account in the optimization of the energy consumption of the extruder system.

[0090] The adaptive digital database of the extruder system is implemented as a digital library, and the repository storage unit has a network interface, which provides access to structured data records via a data transmission network for selection and / or adjustment and / or generation of structured data records, thereby ensuring data exchange and data tracking of the sent, received and generated data through the digital controller and the repository memory. Structuring the data records enables the digital database to search and filter the data records according to structural criteria.

[0091] The digital controller includes a machine learning unit that monitors and classifies input parameter value patterns and adjusts operating setting parameter values ​​of operating units of the feeder and / or extruder and / or forming opening (die) and / or collecting device to align measured material property parameter values ​​of the extrudate within a predetermined tolerance range, thereby enabling the digital controller to build a model for the extruder system based on sample data called training data to make predictions based on input parameter values ​​of the operating setting parameters to extrude extrudates with material property parameters within a predetermined tolerance range of predetermined target parameter values ​​without being constrained by a linear relationship between the input parameters and the operating setting parameters.

[0092] The machine learning unit includes at least one deep learning (DL) structure, which includes one or more neural network (NN) structures and / or one or more statistical modeling structures, which provide output parameter values ​​based on input parameter values ​​indicating parameter value adjustments required to align measured material property parameter values ​​of the extrudate within a predetermined tolerance range to ensure that the machine learning model is a nonlinear function, a piecewise linear function or a step function.

[0093] The deep learning structure based on machine learning includes at least a cascade of multiple layers of nonlinear processing units for feature extraction and signal conversion, and wherein each successive layer uses the output of the previous layer as input, thereby providing supervised learning for at least classification and / or unsupervised learning for at least pattern recognition, thereby ensuring accurate modeling of complex nonlinear models. Increasing computing power enables efficient training of several dependent layers. If the input data is collected or generated by simulation through examples, preferred supervised learning is applied. Another preferred option is to apply an unsupervised learning model, which, by contrast, discovers the inherent structure of unlabeled input parameters on its own.

[0094] The extrusion process of the extruder system is autonomously adjusted by the machine learning unit by automatically adjusting the operating set parameter values ​​of the operating units of the feeder and / or extruder and / or the forming opening and / or the collecting device, so as to align the measured material characteristic parameter values ​​of the extrudate within a predetermined tolerance range by time-based monitoring of the input measured parameter values, thereby ensuring that the operating units can be controllably manipulated in the most efficient manner by the digital controller taking into account the input parameter values.

[0095] Failures of the extruder system during the extrusion process are automatically detected by the machine learning unit based on the measured and monitored input parameter values, wherein an alarm signal and / or a manipulation signal are generated upon detection of a predicted failure within the extrusion process to ensure a controlled extrusion process that produces an extrudate according to predetermined material property parameters. Furthermore, due to the digital controller's intermediate interaction through the alarm signal and / or the manipulation signal, the extrusion system can extrude the extrudate using the input material most efficiently. Furthermore, energy consumption for detecting predicted failures within the extrusion process is inefficient.

[0096] The deep learning of the structural extruder system includes at least a convolutional neural network (CNN) structure as a deep neural network, ensuring local spatial consistency in the input parameters, allowing for fewer weights when sharing certain parameters. This process in the form of a convolution is particularly suitable for extracting information related to the extrusion process, in particular electromagnetic-based data, including optical data generated by a near-infrared sensing unit, with input parameters at low computational cost.

[0097] The measured and / or captured input parameter value patterns are classified and selected with the aid of convolutional layers and pooling layers of a CNN structure, the pooling layers reducing the dimensionality of feature maps of the measured and / or captured input parameter value patterns and thereby reducing the processing complexity within the machine learning unit to adjust the operational setting parameter values ​​of the operational unit.

[0098] The repository storage unit of the extruder system also includes an operation data storage for storing historical operation data, wherein the historical operation data includes historical input parameter values, historical material property parameter values, and predetermined target parameter values, and the machine learning unit is trained by applying the historical operation data to ensure that supervised training using the historical operation data is performed so that the most effective modeling of the deep learning structure is most effectively adapted to the historical behavior of the extruder system characterized by the historical operation data. In addition, the centralized database that communicates / exchanges data with multiple extruder systems increases the data used to train the centralized deep learning structure, which can be accessed (read) by at least one extruder system via the data transmission network and applied by at least one extruder system.

[0099] Material property parameters include texture and / or density and / or color and / or anisotropy and / or chemical composition and / or thickness and / or degree of polymerization and / or moisture content and / or protein content and / or starch content and / or fiber content and / or particle size and / or surface structure and / or tolerance range to ensure the most adequate characterization of the extrudate produced by the extrusion system.

[0100] The operating setting parameters of the extrusion system include the screw speed of the extruder's screw pressing the input material and / or the addition rate of at least one ingredient for constituting the input material through the feeder and / or the adjustment setting of the extruder's regulator for cooling or heating the input material in the extruder and / or the adjustment setting of the regulator of the forming opening and / or the positioning dimensions for the forming opening area, so as to ensure that the operating setting parameter values ​​are effectively generated by the digital controller.

[0101] The target parameters of the extrusion system include at least one of material property parameters and / or process parameters, the process parameters including at least the energy consumption of the extruder system to ensure the most efficient and sustainable extrusion process for producing the extrudate as defined by the material property parameters and / or process parameters and the energy consumption.

[0102] Material parameter values ​​of the input material and / or batches of the input material are automatically determined by the machine learning unit to adjust the dosing process of the feeder by adjusting operating setting parameters of the feeder.

[0103] An extruder network, comprising two or more extruder systems and a central digital controller, the central digital controller comprising a central repository having at least one adaptive central digital database containing structured data records for storing digital recipes and / or ingredients and / or products, and wherein at least one of the digital controllers of the plurality of extruder systems is given read and / or write access to the structured central data records via a data transmission network for selecting and / or adjusting and / or generating structured central data records, and wherein the central data records generated by the extruder systems have at least one data classification parameter, the at least one data classification parameter characterizing the factors that generate the extruder system or influence the generated extruder system, ensuring that the centralized and structured data collection of recipes and / or ingredients and / or products can be accessed by one or more extruder systems via the data transmission network. In addition, the central repository is capable of providing the know-how generated by the extruder systems via the central structured data records. In addition, the centralized digital controller enables the trained centralized deep learning structure to have structured data records of several extrusion systems independent of their geographical location and environmental conditions.

[0104] The data classification parameters of the extruder network include the operating country of the generating extruder system and / or the operator of the generating extruder system and / or the identification of the generating extruder system and / or the extruder type, so as to classify the data records centrally stored in at least one central database using the data identifying the generating extruder system, thereby ensuring a deeper structure of the central data records for multi-dimensional dimensions, so as to enable more effective training of at least one deep learning structure for generating operating setting parameters based on the input parameters of the application. BRIEF DESCRIPTION OF THE DRAWINGS

[0105] The invention will be explained in more detail by way of example with reference to the accompanying drawings, in which:

[0106] Figure 1 A diagram schematically illustrating an exemplary simple schematic diagram of an extrusion process showing the conversion of a raw material, such as a grain meal or starch, into a finished product is shown.

[0107] Figure 2 A block diagram schematically illustrating an exemplary flow chart of an extrusion processing line is shown. The extrusion process begins with characterizing and receiving the ingredients. The ingredients used are critical to the consistency of the extrudate at the end of the processing line. The ingredients are then mixed and / or pretreated, which can be accomplished with equipment such as ribbon blenders and preprocessors to ensure uniformity of the ingredients as they enter the extruder. However, for some products, mixing and pretreatment are optional. Extrusion processing, which is the main cooking step, is then carried out, in which the raw materials are converted into cooked and shaped products. Post-extrusion processing operations such as cutting the extruded product into appropriate sizes, drying the product to the desired moisture, and seasoning or coating to provide the desired flavor and taste are before the product continues to be packaged. In addition to these main processing steps, the convolutional neural network CNN has its intended use.

[0108] Figure 3 A block diagram schematically illustrating an exemplary extrusion system as a multiple-input and multiple-output (MIMO) system is shown. Figure 3 Various inputs and outputs associated with the extruder system of the present invention are shown.

[0109] Figure 4 A diagram illustrating an example of a digital controller including an interface to programmable logic is shown.

[0110] Figure 5 and Figure 6 A description of the input parameters and numerical recipes is provided.

[0111] Figure 7 A diagram schematically illustrating an example of a neural network is shown.

[0112] Figure 8 A diagram illustrating an example of a convolutional neural network including a feature extraction layer and a classification layer is shown.

[0113] Fig. 9 A diagram illustrating an example of an extrusion network including several extruder systems communicating via a data transmission network is shown.

[0114] Fig.10 A diagram illustrating an exemplary simple schematic of an extrusion process showing the conversion of a raw material, such as a grain meal or starch, into a finished product is shown.

[0115] Fig.11 A block diagram schematically illustrating an exemplary process diagram of an extrusion process line is shown. The extrusion process begins with characterizing and receiving the raw materials. The raw materials used are critical to the consistency of the product at the end of the processing line. The raw materials are then mixed and / or pretreated, which can be accomplished with equipment such as ribbon blenders and pre-processors to ensure uniformity of the raw materials as they enter the extruder. However, for some products, mixing and pre-treatment are optional. Extrusion processing, which is the main cooking step, is then carried out, in which the raw materials are converted into cooked and shaped products. Post-extrusion processing operations such as cutting the extruded product into appropriate sizes, drying the product to the desired moisture, and seasoning or coating to provide the desired flavor and taste are before the product is packaged. In addition to these main processing steps, there may be additional steps depending on the type of product produced and its intended use.

[0116] Fig.12 A block diagram schematically illustrating an exemplary extrusion process as a multiple-input and multiple-output (MIMO) system is shown. Figure 3 Various inputs and outputs associated with the extruder system of the present invention are shown. The various extrusion processing parameters can be roughly divided into three categories: (1) independent parameters (input parameters), (2) system parameters (dependent parameters) and (3) product properties (output parameters).

[0117] Fig.13 a / b show diagrams schematically illustrating an exemplary (a) thermoplastic single screw extruder, (b) thermoplastic single screw.

[0118] Fig.14 A description of screw terminology is provided using single screw extruder screw elements.

[0119] Fig.15 A diagram schematically illustrating exemplary co-rotating and counter-rotating intermeshing screws is shown.

[0120] Fig.16A diagram schematically illustrating an exemplary intermeshing co-rotating double-flight screw and 45° kneading blocks is shown.

[0121] Fig.17 A diagram schematically illustrating an exemplary co-rotating screw showing open channels in the intermeshing region is shown.

[0122] Fig.18 A diagram schematically illustrating the general flow pattern in an exemplary co-rotating twin-screw extruder and a counter-rotating twin-screw extruder (in both cases having two flighted screws) is shown.

[0123] Fig.19 Graphs schematically illustrating exemplary shear stress distributions in counter-rotating twin-screw extruders and co-rotating twin-screw extruders are shown.

[0124] Fig. 20 A diagram schematically illustrating an exemplary possible arrangement of co-rotating twin-screw extruders for a multi-process food extrusion system is shown.

[0125] Fig.21 A diagram schematically illustrating an example of expansion of an extrudate at a die exit is shown.

[0126] Fig. 22 A diagram schematically illustrating an example of expansion of a product after leaving a die due to steam pressure differences is shown.

[0127] Fig.23 A diagram schematically illustrating an example of intermeshing orifice plugs (disks) for tight screw confinement and dispersive mixing is shown.

[0128] Fig.24 A diagram schematically illustrating an exemplary twin-screw extruder showing simulated pressure and temperature profiles within the extruder is shown.

[0129] Fig.25 A diagram schematically illustrating an exemplary screw design with two exhaust ports for degassing of volatiles is shown.

[0130] Fig.26 A diagram schematically illustrating an exemplary exhaust port adapter design is shown.

[0131] Fig. 27 A diagram schematically illustrates an exemplary extrusion process implemented by a typical continuous thermomechanical process technology, which combines several unit operations such as conveying, mixing, shearing, plasticizing, melting, cooking and polymerization. These numerous unit operations lead to complex correlations between several variable parameters and their corresponding process responses, and therefore determine product quality.

[0132] Fig.28 A diagram schematically illustrating exemplary AI-based or ML-based automatic recipe management and optimization in the extrusion process of the present invention and the extruder system of the present invention is shown.

[0133] Fig.29 A diagram schematically illustrating an exemplary intelligent recipe selection is shown, where intelligent recipe selection is the process of selecting a suitable recipe from a set of predetermined recipes based on a given product descriptor. A descriptor is defined as a collection of different characteristics of a product. Example characteristics would be color, fiber content, etc. Such a collection would be nameable (in the case of high moisture extrusion) such as "chicken" or "fish", etc.

[0134] Fig.30 A diagram schematically illustrating an exemplary RDB is shown, where the RDB is a database maintained by the system of the present invention that can be used by an operator to search / download new recipes. The database serves as a memory for automatic recipe selection, and the database verifies and updates recipe data by inserting insights from recorded process runs. The database is based on ingredient characteristics, such as color, fiber content, etc., and maps to achievable end product characteristics.

[0135] Fig.31 A diagram schematically illustrating an exemplary IRPC system that generates data for raw materials used in an extrusion process is shown. The IRPC system is intended to serve as: (a) input data for automatic recipe selection by mapping the type of raw materials to the customer / operator's desired product based on the product characteristics that are likely to be achievable. The IRPC system acts as a consultant: the customer / operator can select a specific recipe and the IIC will return whether the desired end product can be achieved using the given ingredients, or it will itself make recommendations on which ingredients should be used in the end product to achieve the best results, and (b) for providing input to intelligent process optimization systems through predicted / anticipated raw product data.

[0136] Fig.32 A diagram schematically illustrates an exemplary ingredient database of the system of the present invention, the ingredient database comprising entries for raw product characteristics such as moisture content, protein content, etc.; and process target information such as CO 2 Correlations, etc., process target information can be contrasted with the measurement results of the final product classification by using product characteristics as classification.

[0137] Fig.33A diagram of an exemplary smart process optimization schematically illustrating the behavior of adjusting parameters of an extrusion process during production based on given targets, in which the targets are process targets (PT) and product targets (given by the recipe). The optimized process must reach or exceed the targets given by the operator. The process optimizer is limited in its optimization by the deviation from the product target, in which the target is the machine recipe or the product characteristic. The process target (PT) appears in the form of a pre-designed command chain.

[0138] Fig.34 A diagram schematically illustrates an exemplary intelligent final product classifier that can, for example, use sensory information retrieved from the final product coming out of the extruder to map the product to final product characteristics. As information sources, the final product classifier can, for example, use data from the process itself (temperature, pressure, SME, etc.) and online final product measurements. To validate its measurements and train the system, offline final product measurements can, for example, be used.

[0139] Fig.35 A diagram schematically illustrating exemplary process objectives is shown. A process objective is a set of process commands that affect how a process operates. Possible process objectives are (i) minimizing the energy consumption of the process, (ii) maximizing the yield of the process, and (iii) maximizing the CO 2 Minimize emissions.

[0140] Fig.36 A diagram schematically illustrating an exemplary automated rework process is shown. Smart rework intake refers to the process of using unusable products generated again from an extrusion process by feeding them back into the process, where the number of reusable products can vary. The service collects its input data from the following sources: the extrusion process itself as well as the final product classification, which yields information about the product status.

[0141] Fig.37 A diagram schematically illustrating an exemplary overall overview of the structure required for automated recipe management and optimization of the present invention is shown.

[0142] Fig.38 An exemplary CO schematically illustrates an embodiment variant of the system according to the invention. 2 eDiagram of the monitoring dashboard.

[0143] Fig.39 A diagram schematically illustrating an exemplary pretreatment process for an exemplary extrusion process is shown.

[0144] Fig.40 A diagram schematically illustrating an exemplary extrusion process is shown. Pretreatment (see Fig.39 ) and extrusion (see Fig.31 ) forms the basis of the entire extrusion process. DETAILED DESCRIPTION

[0145] Figures 1 to 8 The architecture of a possible implementation of an industrial extruder system 1 and an extrusion process of the present invention is schematically illustrated. The industrial extruder system 1 comprises a feeder 101, an extruder 102, a forming opening 103 (die), a collecting device 104 and an extruder control device 500, the feeder 101 feeding a plastically deformable and / or viscous input material 301 to the extruder 102, the extruder 102 continuously pressing the input material 301 to the forming opening 103 and pressing it out from the forming opening according to the operating principle of an Archimedean screw, thereby forming an output material as an extrudate 300. The input material 301 is processed by the extruder 1 by hot extrusion or cold extrusion or warm extrusion or friction extrusion or micro extrusion. The input material 301 extruded by the extruder system 1 may include at least food or metal or polymer or ceramic or concrete or modeling clay. According to Figure 2 The collecting device 104 collects the extrudate 300 for further processing, such as post-extrusion processing including cutting and / or drying and / or seasoning and / or frying and / or coating.

[0146] The extrusion process is controlled in a controlled manner by the extruder control device 500 during operation, and wherein the extruder control device 500 comprises a programmable logic 107 for setting and adjusting the values ​​of the operating setting parameters 208 of the operating unit 105 of the feeder 101, the extruder 102, the forming opening (die) 103 and the collecting device 104. The programmable logic 107 comprises a programmable logic controller of the operating unit, for example a programmable logic controller of a frequency converter for controlling the motor 108. The operating setting parameters 201 comprise the screw speed of the screw of the extruder pressing the input material 301 and / or the addition rate of at least one ingredient constituting the input material 301 through the feeder and / or the adjustment settings of the regulator unit 122 of the extruder 102 for cooling or heating the input material 301 in the extruder 102 and / or the adjustment settings of the regulator of the forming opening 103 and / or the positioning dimensions for the area of ​​the forming opening 103.

[0147] The extruder control device 500 further comprises a digital controller 502 for sending signals to the programmable logic 501 and for manipulating the programmable logic 501, wherein, in order to manipulate the programmable logic 501, the digital controller 502 captures input parameter values ​​200 including at least process parameter values ​​and / or operating setting parameter values ​​208 and / or material property parameter values ​​202 and / or environmental measurement parameter values ​​209. The material property parameters 202 include texture and / or density and / or color and / or anisotropy and / or chemical composition and / or thickness and / or degree of polymerization and / or moisture content and / or protein content and / or starch content and / or fiber content and / or particle size and / or surface structure and / or tolerance range. In addition, as Figure 3 As illustrated in , the input parameter values ​​200 include sensory parameter values ​​207 measured by sensors 111 associated with the feeder 101 and / or the extruder 102 and / or the forming opening 103 and / or the collecting device 104 .

[0148] The extruder control device 500 comprises a repository storage unit 503 having an adaptive digital database 504, wherein the adaptive digital database 504 stores a plurality of selectable structured data records 507 for storing digital recipes, each of the selectable data records 507 comprising at least material property parameters 202 of the input material 301 and target material property parameters 204 of the extrudate 300 and / or initial operating setting parameters, wherein the initial operating setting parameters give initial settings of the operating setting parameters 201 for the operation 105 of the extruder system 1, wherein the selectable data records 507 are configured to store digital recipes. Figure 6 As shown, the input parameter values ​​200 also include the parameter values ​​of the selected data record 507. Figure 3 As shown in the figure, the extruder system 1 includes digital signals for operating the programmable logic 501 and the associated operating unit 507 by means of a digital controller 502 so as to controllably and manipulately extrude an extrudate 300 having a material property parameter value 202 within a predetermined tolerance range of a predetermined target parameter value 206. The digital database 504 is implemented as a digital library, and the repository storage unit 503 has a network interface 506, which provides access to structured data records 507 via a data transmission network 505 for selection and / or adjustment and / or generation of structured data records 507. Data transmission refers to the transmission and reception of data in the form of a digital bit stream or a digitized analog signal transmitted through a point-to-point or point-to-multipoint communication channel. Examples of such channels are copper wires, optical fibers, wireless communications using the radio spectrum, storage media, and computer buses.

[0149] Furthermore, the digital controller 502 includes a machine learning unit 510 that monitors and classifies the input parameter value 200 pattern and adjusts the operating set parameter value 208 of the operating unit 105 of the feeder 101 and / or the operating unit 105 of the extruder 102 and / or the operating unit 105 of the forming opening 103 and / or the operating unit 105 of the collecting device 104 to align the measured material property parameter value 202 of the extrudate 300 within a predetermined tolerance range. The machine learning unit 510 can be implemented, for example, by including a deep learning (DL) structure, which includes, for example: Figure 7 One or more neural network (NN) structures illustrated in FIG. 6 , one or more neural network (NN) structures may, for example, include an input layer 601, at least one hidden layer 602, and an output layer 603; and / or one or more statistical modeling structures, one or more statistical modeling structures providing output parameter values ​​213 based on input parameter values ​​200 indicating parameter value adjustments required to align measured material property parameter values ​​202 of extrudate 300 within a predetermined tolerance range. The machine learning-based DL structure includes at least a cascade of multiple layers of nonlinear processing units 512 for feature extraction 512a and signal transformation 512b, and wherein each successive layer uses the output of the previous layer as input, thereby providing supervised learning for at least classification and / or unsupervised learning for at least pattern recognition. In data mining, supervised learning is divided into two types of problems: classification and regression: (i) Classification problems use algorithms to accurately assign test data to specific categories. Alternatively, in the real world, supervised learning algorithms can be used to classify spam in separate folders in an inbox. Linear classifiers, support vector machines, decision trees, and random forests are all common types of classification algorithms. Preferably, the DL structure includes at least Figure 8 The structure shown in the figure is a convolutional neural network (CNN) of a deep neural network.

[0150] Regression is another type of supervised learning method that uses algorithms to understand the relationship between the dependent and independent variables. Regression models help predict numerical values ​​based on different data points, such as sales revenue forecasts for a given business. Some popular regression algorithms are linear regression, logistic regression, and polynomial regression. Unsupervised learning uses machine learning algorithms to analyze and cluster unlabeled data sets. These algorithms find hidden patterns in the data without human intervention, so these algorithms are "unsupervised". Unsupervised learning models are used for three main tasks: clustering, association, and dimensionality reduction: (i) Clustering is a data mining technique used to group unlabeled data based on their similarities or differences. For example, the K-means clustering algorithm assigns similar data points into groups, where the K value represents the size and granularity of the grouping. This technique helps in market segmentation, image compression, etc.; (ii) Association is another type of unsupervised learning method that uses different rules to find the relationship between variables in a given data set. These methods are often used in shopping basket analysis and recommendation engines, similar to "customers who bought this item also bought" recommendations; (iii) Dimensionality reduction is a learning technique used when the number of features or dimensions in a given data set is too high. Dimensionality reduction reduces the number of data inputs to a manageable size while also maintaining the integrity of the data. Typically, this technique is used in the pre-processing data stage, such as when an autoencoder removes noise from visual data to improve image quality. The measured and / or captured input parameter value 200 patterns are classified and selected with the help of the convolutional layer 612 and pooling layer 613 of the CNN structure, which reduces the dimensionality of the feature map of the measured and / or captured input parameter value 200 pattern and thereby reduces the processing complexity within the machine learning unit 510 to accommodate the operation setting parameter value 201 of the operation unit 105.

[0151] Faults during the extrusion process are automatically detected by the machine learning unit based on the measured and monitored input parameter values ​​200, wherein an alarm signal and / or a manipulation signal is generated when a predicted fault within the extrusion process is detected. Preferably, the digital controller 502 comprises a fault handling unit that generates an error signal based on the alarm signal and / or the manipulation signal. The operating setting parameters comprise an error signal for manipulating the programming logic 501, in particular for setting an error function of a corresponding programming logic controller of a corresponding operating unit 105. The fault handling unit can be configured for a given alarm signal and / or manipulation signal. Preferably, the digital controller comprises a human-machine interface for issuing an alarm signal.

[0152] The extruder network 2 comprises two or more extruder systems 1 and a central digital controller 514, the central digital controller 504 comprising a central repository 512 having at least one adaptive central digital database 513 containing structured data sets 515 for storing digital recipes and / or ingredients and / or products, and wherein at least one of the digital controllers of the plurality of extruder systems 1 is given read and / or write access to the structured central data records 515 via the data transmission network 505 for selecting and / or adjusting and / or generating the structured central data records 515, and wherein the central data records generated by the extruder system 1 have at least one data classification parameter 211 characterizing the generating extruder system or a factor influencing the generating extruder system 1. The data classification parameter 211 comprises the operating country of the generating extruder system and / or the operator of the generating extruder system 1 and / or the identification of the generating extruder system 1 and / or the extruder type, so as to classify the data records centrally stored in the at least one central database using the data classification.

[0153] Ⅰ. Consistent product quality and unmanned operation

[0154] Extrusion is a continuous thermomechanical process technology that combines several unit operations such as conveying, mixing, shearing, plasticizing, melting, cooking and polymerization. These numerous unit operations lead to complex correlations between several variable parameters and their corresponding process responses, and thus determine the product quality (see Fig.18 ). The current situation requires time-intensive and cost-intensive off-line analysis, well-educated operators and laboratory personnel to assess product quality and adjust the process accordingly and thus adjust product quality. Due to the nature of off-line analysis, process adjustments are completed with a delay of some time. Therefore, online measurements that assess product quality in real time will allow for immediate process adjustments and consistent product quality. In addition, off-line analysis will be reduced, thereby reducing both analysis and personnel costs.

[0155] The settings of the extruder and its peripherals take into account fixed parameters such as screw configuration, barrel length, cooling die geometry and cooling die inserts. Variable parameters are temperature, screw speed, moisture addition rate, solid addition rate, oil addition rate and the location of liquid addition such as water, oil, fragrance, mineral solution, acid, caustic. In addition to this, nitrogen (N2) introduced as part of the aeration technology is also a variable parameter. Variable parameters determine process responses such as specific mechanical energy (SME), thermal stress (STE), pressure at the end plate, temperature at the end plate, residence time (distribution), weighted average total strain (WATS, Forte) and flow distribution / velocity at the outlet of the cooling die. All the process responses mentioned above are measurable quantities. SME is a quantity calculated from power consumption measurements / torque measurements recorded by the main drive of the extruder, STE is a quantity calculated based on the inlet and outlet temperatures, and the volume flow of the heating cycle, the pressure and temperature at the end plate are quantities measured by pressure and temperature probes, the flow distribution and velocity can be obtained by a high-definition camera and an underlying video analysis program, the residence time (distribution) is a quantity measured by using an inert indicator, such as a colorant, which is then also recorded by a high-definition camera and tracked over time, and WATS (weighted average total strain) is a quantity calculated using fixed parameters (screw configuration, barrel length, cooling die geometry and inserts), screw speed and residence time. As an embodiment variant, the ML-based or AI-based core engine 11 includes an AI-controlled loop to change the process response by adjusting variable parameters. In addition, a complex interaction between process response and variable parameters and extrudate product quality can be established. Since a change in one single variable parameter may have more than one effect on the process response and thus affect the product quality without a clear correlation, the artificial intelligence achieved by the present invention is a powerful tool to reveal these interactions, not only for daily production and quality assurance, but also for R&D purposes. As an example, there are prior art systems that aim to measure and study the effect of SME on the anisotropy of the final product, however, it turns out that when increasing the screw speed to change the SME, the temperature at the end plate also increases. Therefore, with the prior art system, it is not possible to clearly link the observed effects on product quality.

[0156] The product quality of the wet extrudate is mainly described by its texture, anisotropy, density, chemical composition, color and degree of polymerization. In order to establish a manually controlled loop to adjust the variable parameters in real time, the product quality must be analyzed online. The following table (Table 2 is a non-exhaustive list) shows the applicable methods to achieve consistent product quality throughout the production process.

[0157]

[0158]

[0159] Table 2 shows examples of product qualities and applicable online methods.

[0160] If some of the above product qualities cannot be measured directly, it is necessary to implement a mapping between measurable process parameters such as temperature, pressure and SME and between the above method and the product quality indicators. As an implementation variant, a machine learning modeling structure is used, for example, to perform appropriate mapping.

[0161] The resulting final product quality indicators can be used for a variety of different applications within the extrusion process, for example as validation data for AI applications previously used in the process chain, such as automated process optimization, where process goals are defined by the operator, and instead of using conventional recipes that can then decide between good and bad products, ingredients can be mapped to achievable products in an ingredient database.

[0162] II. Automated recipe management and optimization

[0163] With regard to automated recipe management and optimization, there are various technical problems that are solved by the extruder and extrusion system of the present invention: (i) in prior art systems, it is often unclear whether the maximum potential of any extrusion machine has been reached; (ii) in prior art systems, operators often do not know much about how to operate, and customers also often do not know much about how to operate; (iii) in prior art systems, optimizing recipes is often very time-consuming. All happens in the minds of technical experts, i.e., based on experience; (iv) in prior art systems, optimized product characteristics are often unclear and need to be determined through an approximation process based on experience; and (v) in prior art systems, recipe management and optimization still need to be embedded in the extrusion process, i.e., form part of the standard extrusion process. Fig.28 A schematic diagram schematically illustrating an exemplary overall overview of an architecture suitable for implementing the automated recipe management and optimization of the present invention is shown.

[0164] (a) Intelligent recipe selection

[0165] Intelligent recipe selection is the process of selecting a suitable recipe from a set of predetermined recipes based on a given product descriptor (see Fig. 20 ). Descriptors are defined as a collection of different characteristics of a product. Exemplary characteristics would be color, fiber content, etc. Such a collection would be nameable (in case of high moisture extrusion), like "chicken" or "fish", etc. Depending on the collection, the included characteristics can be changed by the operator after selection. For example: the operator changes the color of the final product.

[0166] The recipe selection process is based on different data sources:

[0167] Recipe database RDB

[0168] Intelligent ingredient classification IIC

[0169] Intelligent Final Product Classification IFPC (described in a later section)

[0170] Process Objective PT (described in a later section)

[0171] Operator's choice

[0172] Recipe Database (RDB)

[0173] The RDB is a database maintained by the system of the present invention that can be used by operators to search / download new recipes (see Fig.21 ). The database is used as a memory for automatic recipe selection, the database verifies and updates recipe data by inserting insights from recorded process runs. The database is based on ingredient characteristics, such as color, fiber content, etc., and is mapped to achievable end-product characteristics.

[0174] Ingredient Classification (IIC)

[0175] The IRPC system generates data on raw materials used in the extrusion process. Its objectives are to:

[0176] Input data for automatic selection of recipes by mapping the types of raw materials to the customer / operator desired product based on the product characteristics that are likely to be achievable. The IRPC system acts as a consultant: the customer / operator can select a specific recipe and the IIC will return whether the desired end product can be achieved using the given ingredients, or it will make its own recommendations on which ingredients should be used in the end product for best results.

[0177] Used to provide input to smart process optimization systems through predicted / anticipated raw product data.

[0178] IRPC systems are available from the following sources:

[0179] • Ingredient database IDB (see below).

[0180] Online raw product analysis through online equipment such as NIR.

[0181] Offline raw product analysis, such as laboratory

[0182] · Return the final product classification IFPC of the observed final product characteristics

[0183] Process goals selected by the operator

[0184] Operator manually enters original product information

[0185] IRPC uses measured online raw product analysis data and offline raw product analysis data to detect deviations in the product. These deviations are stored in the raw material database and used to train and validate the IRPC model and its input data from the database. IRPC sends its predicted raw product characteristics to the automatic process optimization to influence the process (see also Fig. 22 ).

[0186] Ingredient Database (IDB)

[0187] The ingredient database maintained by the system of the present invention contains entries for raw product characteristics such as moisture, protein content, etc.; contains process target information, such as CO 2 Correlations, etc., process target information can be contrasted with the measurement results of the final product classification by using product characteristics as classification.

[0188] Smart Process Optimization (SPO)

[0189] Intelligent process optimization is the act of adjusting the parameters of the extrusion process during production based on given targets, in which the targets are process targets (PT) and product targets (given by the recipe). The optimized process must reach or exceed the targets given by the operator. The process optimizer is limited in its optimization by the deviation of the product target, in which the target is the machine recipe or product characteristics. The process target (PT) appears in the form of a pre-designed command chain.

[0190] SPO uses the following data sources:

[0191] • Intelligent Final Product Classifier (IFPC) returns measured final product characteristics leaving the process.

[0192] • The extrusion process generates real-time data from the process, like product temperature at a given position in the process chain, pressure at the end plate and pressure at the (cooling) die, SME etc. If recipe deviations are used, these values ​​will be used as deviation limits.

[0193] The Smart Recipe Selector (SMR) sends the final product characteristics to be achieved.

[0194] Intelligent Ingredient Classifier (IIS) sends its predicted raw product characteristics information

[0195] Online product measurement / offline product measurement to send real-time batching information / batch batching information

[0196] Operator selects process target (PT)

[0197] • Input data of possible rework to feed into the intelligent rework intake in process.

[0198] The Smart Process Optimizer can operate in two different modes: Assuming that both the Smart Recipe Selector and the Final Product Classifier are available, the Smart Process Optimizer can operate in the Characterized Product Mode, in which it controls the process based on the product characteristics to be achieved. Assuming that no Smart Recipe Selector is available or no Final Product Classifier is available, the Smart Process Optimizer can operate in the actual Data-only Mode, in which it optimizes based on online measurements.

[0199] Intelligent Final Product Classification (IFPC)

[0200] The intelligent final product classifier uses sensory information retrieved from the final product coming out of the extruder to map the product to final product characteristics. As information sources, the final product classifier can use, for example, data from the process itself (temperature, pressure, SME, etc.) and online final product measurements. In order to validate its measurements and train the system, offline final product measurements can be used, for example.

[0201] Its output data, which reflects the characteristics of the final product, is reused by multiple systems:

[0202] • Intelligent process optimizers that use this data to adjust the process. The measurements are used as quality metrics in trying to achieve target product characteristics.

[0203] Smart rework intake process, where analysis of unusable products is used to analyze how much product can be immediately fed back into the process to reduce waste

[0204] • An intelligent ingredient classifier that uses the output of the IFPC to map ingredients to achievable end products.

[0205] Process Target (PT)

[0206] A process goal is a collection of process commands that affect how a process runs (see Fig.26 ). Possible process objectives include:

[0207] Minimize the energy consumption of the process

[0208] Maximize the yield of the process

[0209] · Make CO 2 Minimizing emissions

[0210] Process objectives serve as input to the following systems:

[0211] · Intelligent batching classifiers, where the intelligent batching classifier has an impact on the advisory function based on the target. For example, if the target CO is selected 2emissions, the ingredients will be based on their CO 2 Select based on relevance.

[0212] Intelligent recipe selector, where recipes are proposed and selected based on the achievement of their specific goals

[0213] Intelligent process optimizer, where the chain of command has an impact on the production process itself.

[0214] Automatic rework

[0215] Smart rework intake is a process that uses unusable products generated from the extrusion process again by feeding them back into the process, where the amount of product that can be reused can vary (see Fig. 27 ). The service collects its input data from the following sources: the extrusion process itself and the final product classification, which yields information about the product status. Whether a product can be fed back into the extrusion process depends on the following parameters:

[0216] ·Set product characteristic values ​​as targets. Products that do not meet the targets can be reused.

[0217] Product moisture. If the product is too wet, it will not be possible to return the product without it being sticky.

[0218] Product block size. Large pieces will not be able to be shipped.

[0219] Rework intake will have an impact on Smart Process Optimization. This product cannot be understood as a normal "ingredient" for the process. Smart Process Optimization will receive recommendations from the Smart Rework Intake Controller on the amount of rework to ingest.

[0220] III. Ingredient database (cloud) for process optimization based on historical data

[0221] Extrusion allows a wide range of different ingredients to be processed. The ingredients and their properties depend on several environmental factors, but also on the process steps taken during their production. Ingredients are therefore susceptible to variations in their composition and functionality.

[0222] For high-moisture extruded meat-like products, different protein sources are usually used, such as flour (protein about 50%), concentrates (protein 60%-80%) and isolates (protein>80%), or even intermediate products from their production process such as slurries. In an embodiment variant, in order to enhance the intelligent extruder system, an ingredient database is created to continuously optimize the extrusion process. In addition to different protein sources and water, the recipe also includes carbohydrates (i.e. starch, fiber, sugar), fats / oils, minerals, salts, acids, corrosives, flavorings. A comprehensive ingredient database must contain the composition and function-related properties of the corresponding ingredients.

[0223] First, the ingredients are analyzed to determine their chemical composition, namely solids content (moisture content), protein content, fat content, carbohydrate content and minerals.

[0224] Furthermore, specific knowledge about the protein composition of the ingredients is very important as it determines the potential for later texturization capabilities during the process. More specifically, the intermolecular and intramolecular interactions during texturization are a function of the specific amino acid profile. During high moisture extrusion, most interactions between proteins are non-covalent interactions such as hydrogen bonds, hydrophobic interactions, van der Waals interactions, and electrostatic interactions. Furthermore, sulfur-rich amino acids, especially cysteine, form covalent disulfide bonds, which are the main interactions for the formation of heat-stable high moisture extrudates. Therefore, a detailed analysis of the ingredients is required. In addition to the knowledge about the amino acid profile, advanced knowledge such as denaturation temperature and kinetics as well as aggregation temperature and polymerization kinetics are also part of the ingredient database. Therefore, the ingredient database helps in formulation development and provides insights into whether the ingredient participates in texturization or serves as an inert filler in the final product.

[0225] Further analysis may be required to identify specific carbohydrates to differentiate between sugars, fibers, and starches. For starch-rich ingredients, it may be important to include knowledge about the amylose-amylopectin ratio and gelatinization kinetics and degradation. For sugar-rich ingredients, it may be important to include knowledge about melting temperatures and caramelization.

[0226] In general, the characteristics of the ingredients such as true density, particle size distribution, fluidity, wettability, dispersibility, water absorption, ionic strength and pH value of the slurry, viscosity, elastic modulus, phase transition from dispersion (viscoelastic liquid) to plasticized substance (viscoelastic solid), as well as the time, temperature and strain dependence of all these parameters are necessary for the ingredient database. The ingredient database is a continuously growing database that helps in the evaluation of new ingredients and the development of new recipes. In addition, variable parameters for the optimization of the extrusion process can be derived.

[0227] Considering the above list of required fluid and solid properties and the lack of detailed understanding of cooking, kneading and conveying, it is unlikely that a physics-based model of the entire extrusion process will be available anytime soon. However, significant attempts are currently being made to model the subsystems of the extrusion process, particularly die flow of wet texturized proteins.

[0228] Quantification of many of the above mentioned properties is a challenge in itself, as laboratory equipment often cannot simulate the real process, and online measurements are inaccurate because the available sensors are insufficient or inappropriately located. Often, only pressure and surface temperature are available when viscosity and density values ​​are needed to provide information for the material model.

[0229] Modeling viscoelastic flows in industrial processes with complex and interlocking geometries is extremely difficult and time-consuming, but in principle possible. Including phase changes of complex fluids and semi-solids requires innovative simulation techniques, such as meshless methods. This is not yet an industry standard and is also an area of ​​ongoing academic research. In the prior art, three software codes are known to have been used in similar challenges or are able to capture the main physical characteristics of such processes.

[0230] However, such simplified modeling approaches to the extrusion process can only achieve limited success. From today's perspective, creating material models and measuring the corresponding material properties would require a lot of resources and time to reach a level of industrial usability. Therefore, data-based methods offer an alternative and more accessible way to (try to) relate input parameters to result quantities and build a "model" that will ultimately allow prediction and optimization of the process.

[0231] IV. Optimized and fully automated startup time

[0232] The current situation for the extrusion process requires each operator to carry out a separate startup procedure, resulting in a waste of raw materials and time (about 30 minutes) because the startup phase is only objectively evaluated. Therefore, the intelligent extruder system of the present invention enables a more accurate and optimized startup phase when taking into account the measured quantities for online evaluation of the texture (e.g. the cutting force at the die) and the process response (e.g. the pressure at the end plate). For example, the pressure probe at the end plate (transition from the extruder barrel to the cooling die) is a suitable number to rely on for the automation of the startup process. When the pressure at the end plate rises to above 2 bars, it indicates that the screw section is filled. Therefore, in this embodiment variant, the production phase is started automatically.

[0233] In addition, the slurry is measured online during the startup phase (e.g., chemical composition by NIR), making it possible to use it for rework. For example, as an embodiment variant, the rework is injected into the barrel. In this case, the intelligent extruder system relies on the ingredient database to adjust the variable parameters and ultimately maintain a consistent product quality during production.

[0234] V.Safety in production

[0235] It is known in the prior art that plant-based meat analogs are more perishable and more susceptible to spoilage than meat due to their high protein and moisture contents and neutral pH, and therefore require more attention in meat analog production. In particular, an effective food safety management system is essential to ensure safe production.

[0236] Online / Inline CCP Monitoring

[0237] Critical Control Points (CCP):

[0238] Temperature in the end plate

[0239] A temperature sensor is installed in the end plate to record the temperature every 1 second. The real-time temperature curve will be displayed on the dashboard, which has the following functions:

[0240] Calculate Log5 reduction conditions based on recipe

[0241] Set critical values ​​for key food safety parameters (temperature, duration, etc.) based on Log5 reduction conditions

[0242] Alarm and warning functions when the temperature falls below a critical threshold (CCP error)

[0243] Displays an alarm summary and a list of all time windows / time periods when a CCP error occurs

[0244] Each alarm and CCP error can be selected for further analysis

[0245] Ability to track product errors and see what happened to the product and what actions were taken (e.g. equipment was cleaned and sanitized, etc.)

[0246] Generate reports showing CCPs, critical limits, CCP errors, recipes, time frames, operators, corrective actions, etc.

[0247] Online / Inline CCP Control

[0248] When a CCP error occurs and the operator does not take corrective action, the control system will automatically adjust the process based on the recipe to increase the temperature above the critical threshold.

[0249] A guide baffle is installed at the outlet of the cooling mold, and the products produced during the period that does not meet the CCP standard will go to the waste box

[0250] Verification of machine cleanliness

[0251] In the preprocessor (see Fig.30 ), extruder (see Fig.31), cooling dies, conveyors and cutting machines. After cleaning, optical sensors are used to assess cleanliness (by detecting residues and microbial contamination, etc.), and an alarm system is established to warn when cleanliness does not meet standards and corrective action is required. For wet cleaning, visual inspection is usually not enough. Microbiological testing should be performed. ATP testing can be recommended (quick test: 5 minutes). A checkbox is created on the dashboard, and the operator must check in the checkbox whether a microbiological test has been performed. Alternatively, machine cleanliness can be determined by indirect methods (such as microbiological testing of downstream products).

[0252] Safety checks before starting the extrusion process

[0253] Before starting the extrusion process, the control system will perform a self-test procedure:

[0254] Check cleaning history (e.g. when was the last time it was cleaned)

[0255] Check microbiological testing (e.g. when was the last test)

[0256] Optical cleanliness inspection

[0257] Based on the feedback from the above mentioned checks, the control system will decide whether the extrusion process can be started or if further cleaning / testing is required.

[0258] VI. Sustainability

[0259] CO 2 Equivalent (CO 2 e) is a standard unit for measuring carbon footprint and is used as a universal indicator due to the fact that it takes into account all greenhouse gases and is easily available. It simplifies the problem of climate change and highlights carbon hotspots so that targeted actions can be taken. In addition, it complies with ISO standards.

[0260] Ingredient database for consulting services

[0261] For all relevant ingredients, their potential suppliers and relevant product specifications, costs and CO 2 The emission coefficients etc. are recorded in the batching database.

[0262] Based on the specific requirements of customers, we can recommend suitable ingredient suppliers

[0263] Depending on the ingredients, suppliers, transportation, etc., CO may be generated. 2 e Summary of emissions.

[0264] Can generate CO with different ingredients, suppliers and transportation 2 e Comparison of emissions.

[0265] Process CO 2 e-monitoring

[0266] When the system is connected to the Expert Insight system, information on batching, production and energy consumption is obtained, which can be used to calculate the CO 2 eEmission quantification.

[0267] Digital monitoring and expert services are available, such as the use of appropriate CO 2 e Monitoring Dashboard (see Fig.29 ):

[0268] Real-time CO of the entire process 2 Emissions

[0269] CO per processing step 2 e Detailed breakdown of emissions (can identify carbon hotspots)

[0270] CO from SGS 2 e-Certification

[0271] Track sustainability performance over time

[0272] The Expert Insight system can be implemented as a central platform for connecting products and services, optimizing the efficiency of the plant and reducing maintenance time, energy consumption and waste. Transforming machinery into connected devices allows data to flow freely from sensors, machines and control units to a single secure storage location. Connecting the plant to a central platform forms a portal and takes advantage of digitalization. The central platform allows transparency to be brought to process and machine data - for example equipped with individual dashboards showing the most important KPIs. This transparency makes it possible to provide or initiate or electronically signal specific actions to improve performance and optimize processes. Thus, connected devices, operating indicators and analytics help optimize the efficiency of the plant. Choose from a wide range of digital services to increase productivity, improve product quality and reduce waste and energy consumption. With access via a tablet or smartphone, this even allows the control of the plant on the move as an implementation variant.

[0273] Reference numerals list

[0274] 1 Extruder system

[0275] 2Distributed extruder network system

[0276] 101 Feeder

[0277] 102 Extruder

[0278] 103 forming opening

[0279] 104 Collection Device

[0280] 105 operation unit

[0281] 111 Sensor

[0282] 120 screw

[0283] 121 barrel

[0284] 122 Adjustment Unit

[0285] 123 Motor

[0286] 200 Input parameters

[0287] 201 Operation Setting Parameters

[0288] 202 Material characteristic parameters

[0289] 203 Environmental Measurement Parameters

[0290] 204 Target material characteristic parameters

[0291] 206 Predetermined target parameter value

[0292] 207 sensor parameters

[0293] 208 Operation Setting Parameters

[0294] 209 Environmental Measurement Parameters

[0295] 211 Initial Operation Setting Parameters

[0296] 212 Digital Recipe

[0297] 213 Output Materials

[0298] 300 extrudate

[0299] 301 Input Materials

[0300] 302 Output Material

[0301] 303 Ingredients

[0302] 401 Action Settings

[0303] 500 Extruder Control Device

[0304] 501 Programmable Logic

[0305] 502 Digital Controller

[0306] 503 Repository Storage Unit

[0307] 504 Digital Database

[0308] 505 Data Transmission Network

[0309] 506 Network Interface

[0310] 507 Data Records

[0311] 511 Operation Data Memory

[0312] 512 nonlinear processing units

[0313] 512 Central Repository

[0314] 513 Central Digital Database

[0315] 514 Central Digital Controller

[0316] 515 operation unit

[0317] 515 Central Data Record

[0318] 601 Input layer

[0319] 602 hidden layers

[0320] 603 Output layer

[0321] 610 Machine Learning Unit

[0322] 611 Deep Learning (DL) Architecture

[0323] 612 convolutional layers

[0324] 613 Pooling layer

Claims

1. An extruder system (1), comprising a feeder (101), an extruder (102), a forming opening (103) (die), a collecting device (104) and an extruder control device (500), wherein the feeder (101) feeds a plastically deformable and / or viscous input material (301) to the extruder (102), and the extruder (102) continuously presses the input material (301) to the forming opening (103) and presses it out from the forming opening (103), thereby forming an output material as an extrudate (300), wherein: The collecting device (104) collects the extrudate (300) for further processing, wherein the extrusion process is controlled in a controllable manner by the extruder control device (500) during operation, and wherein the extruder control device (500) comprises a programmable logic (107) for setting and adjusting the operating setting parameter values ​​(208) of the operating unit (105) of the feeder (101), the extruder (102), the forming opening (die) (103) and the collecting device (104), characterized in that The extruder control device (500) further comprises a digital controller (502) for signaling and manipulating the programmable logic (501), wherein, in order to manipulate the programmable logic (501), the digital controller (502) captures input parameter values ​​(200) comprising at least process parameter values ​​and / or operating setting parameter values ​​(208) and / or material characteristic parameter values ​​(202) and / or environmental measurement parameter values ​​(209), wherein the input parameter values ​​comprise sensory parameter values ​​(207) measured by sensors (111) associated with the feeder (101) and / or the extruder (102) and / or the forming opening (103) and / or the collecting device (104), The extruder control device (500) comprises a repository storage unit (503) having an adaptive digital database (504), wherein the adaptive digital database (504) stores a plurality of selectable structured data records (507) for storing digital recipes, each of the selectable data records (507) comprising at least material property parameters (202) of the input material (200) and target material property parameters (204) of the extrudate (300) and / or initial operating setting parameters giving initial settings of operating setting parameters (201) for the operation (105) of the extruder system (1), wherein the input parameter values ​​(200) further comprise parameter values ​​of the selected data records (507), and The extruder system (1) includes digital signals for manipulating the programmable logic (501) and associated operating units (507) by means of the digital controller (502) so as to controllably and manipulably extrude an extrudate (300) having a material property parameter value (202) within a predetermined tolerance range of a predetermined target parameter value (206).

2. The extruder system (1) according to claim 1, wherein: The adaptive digital database (504) is implemented as a digital library, and the repository storage unit (503) has a network interface (506), which provides access to the structured data records (507) via a data transmission network (505) for selecting and / or adjusting and / or generating the structured data records (507).

3. The extruder system (1) according to claim 1, wherein: The digital controller (502) includes a machine learning unit (510) which monitors and classifies input parameter value (200) patterns and adjusts the operating setting parameter values ​​(208) of the operating unit (105) of the feeder (101) and / or the extruder (102) and / or the forming opening (103) and / or the collecting device (104) to align the measured material property parameter values ​​(202) of the extrudate (300) within the predetermined tolerance range.

4. The extruder system (1) according to claim 3, wherein: The machine learning unit (510) includes at least one deep learning (DL) structure, which includes one or more neural network (NN) structures and / or one or more statistical modeling structures that provide output parameter values ​​() based on the input parameter values ​​(200) indicating the parameter value adjustment required to align the measured material property parameter values ​​of the extrudate (300) within the predetermined tolerance range.

5. The extruder system (1) according to claim 3, wherein: The deep learning structure based on machine learning includes at least a cascade of multiple layers of nonlinear processing units (512) for feature extraction and signal conversion, and wherein each consecutive layer uses the output of the previous layer as input, thereby providing supervised learning at least for classification and / or unsupervised learning at least for pattern recognition.

6. Extruder system (1) according to one of claims 3 to 5, wherein: The extrusion process is autonomously adjusted by the machine learning unit by automatically adjusting the operating set parameter values ​​of the operating units of the feeder and / or the extruder and / or the forming opening and / or the collecting device so as to align the measured material characteristic parameter values ​​of the extrudate within the predetermined tolerance range by time-based monitoring of the input measured parameter values ​​(200).

7. The extruder system according to one of claims 3 to 6, wherein: Faults during the extrusion process are automatically detected by the machine learning unit based on measured and monitored input parameter values ​​(200), wherein an alarm signal and / or a maneuvering signal is generated upon detection of a predicted fault within the extrusion process.

8. Extruder system (1) according to one of claims 4 to 7, wherein: The deep learning structure includes at least a convolutional neural network (CNN) structure as a deep neural network.

9. The extruder system (1) according to claim 8, wherein: The measured and / or captured input parameter value (200) patterns are classified and selected with the aid of the convolutional layer (612) and the pooling layer (613) of the convolutional neural network structure, wherein the pooling layer (613) reduces the dimension of the feature map of the measured and / or captured input parameter value (200) pattern and thereby reduces the processing complexity within the machine learning unit (510) to adjust the operation setting parameter value (201) of the operation unit (105).

10. The extruder system (1) according to one of claims 1 to 9, wherein: The repository storage unit (503) also includes an operation data storage (511) for storing historical operation data, wherein the historical operation data includes historical input parameter values, historical material property parameter values ​​and predetermined target parameter values, and the machine learning unit is trained by applying the historical operation data.

11. The extruder system (1) according to claims 1 to 10, wherein: Material property parameters include texture and / or density and / or color and / or anisotropy and / or chemical composition and / or thickness and / or degree of polymerization and / or moisture content and / or protein content and / or starch content and / or fiber content and / or particle size and / or surface structure and / or tolerance range.

12. The extruder system (1) according to one of claims 1 to 11, wherein: The operating setting parameters (201) include the screw speed of the screw of the extruder pressing the input material, and / or the addition rate of at least one ingredient for constituting the input material through the feeder, and / or the adjustment setting of the regulator of the extruder for cooling or heating the input material in the extruder, and / or the adjustment setting of the regulator of the forming opening, and / or the positioning dimensions for the forming opening area.

13. The extruder system according to one of claims 1 to 12, wherein: The target parameters include at least one of a material property parameter and / or a process parameter including at least the energy consumption of the extruder system.

14. The extruder system of claim 1, wherein: The material parameter values ​​of the input material and / or the dosage of the input material are automatically determined by the machine learning unit to adjust the dosing process of the feeder by adjusting the operating setting parameters of the feeder.

15. A decentralized extruder network system (2) comprising two or more extruder systems (1) according to any one of claims 1 to 14 and a central digital controller (514), the central digital controller (514) comprising a central repository (512) with at least one adaptive central digital database (512), the at least one adaptive central digital database (512) containing structured data sets (515) for storing digital recipes and / or ingredients and / or products, and wherein, At least one of the plurality of digital controllers of the extruder system (1) is given read and / or write access to a structured central data record (515) via a data transmission network (505) for selecting and / or adjusting and / or generating the structured central data record (515), wherein the central data record generated by the extruder system (1) has at least one data classification parameter (211) characterizing the generating extruder system or a factor influencing the generating extruder system (1).

16. The decentralized extruder network system (2) according to claim 15, wherein: The data classification parameters (211) include the country of operation of the production extruder system and / or the operator of the production extruder system (1) and / or the identification and / or extruder type of the production extruder system (1), so as to use the data classification parameters (211) to classify the data records centrally stored in at least one central database.

Images