EP0402940A2 - Method for mixing textile fibres - Google Patents

Abstract

• I wenigstens folgende Angaben in einen Rechner (116) eingegeben werden:
  • a) die Eigenschaften der Fasern der einzelnen Komponenten, und
  • b) die gewünschten Eigenschaften des aus der Faser­mischung hergestellten Kardenbandes bzw. Garnes;
• II aus diesen Vorgaben der Rechner entsprechend einem vorgegebenen Rechenalgorithmus eine Komponentenauf­teilung errechnet, die bei minimierter Abweichung von dem gewünschten Kardenband bzw. den Garneigenschaften diese wenigstens annähernd erfüllt, und ggf. eine Korrektur der errechneten Komponentenaufteilung unter Berücksichtigung etwaiger, ebenfalls in den Rechner eingegebenen Randbedingungen bzw. Sonderwünsche vornimmt und eine korrigierte Komponentenaufteilung errechnet,
• III die vom Rechner ermittelte Komponentenaufteilung bzw. korrigierte Komponentenaufteilung für die Einstellung bzw. Regelung der Zuspeisung der einzelnen Komponenten am Mischer (6) ausgenutzt wird, um die errechnete und ggf. korrigierte Komponentenaufteilung in dem vom Mischer gelieferten Fasergemisch zu erhalten.

A method for mixing textile fibers, in which different fibers of fiber bales (2) of different provenance are removed and mixed, is characterized in that

• I at least the following information is entered into a computer (116):
  • a) the properties of the fibers of the individual components, and
  • b) the desired properties of the card sliver or yarn produced from the fiber mixture;
• II a computation of components is calculated from these specifications of the computer in accordance with a predefined calculation algorithm, which at least approximately fulfills these with a minimal deviation from the desired card sliver or the yarn properties, and, if necessary, a correction of the computation of components calculated taking into account any boundary conditions or parameters also entered into the computer Makes special requests and calculates a correct component allocation,
• III the component division or corrected component division determined by the computer is used for setting or regulating the feeding of the individual components on the mixer (6) in order to obtain the calculated and possibly corrected component division in the fiber mixture supplied by the mixer.

Description

The present invention relates to a process for mixing textile fibers, in which different types of fibers are removed from fiber bales of different provenance and mixed.

The previous methods of mixing consist either in that fiber bales from different origins are placed in a row and removed by means of a removal device which moves back and forth by loosening fiber flakes from the surface and transferring them to a means of transport or in that parts manually or mechanically lifted from fiber bales and fed one after the other on a conveyor belt to an opening machine, in which these parts are broken down into fiber flakes and transferred to a means of transport.

Such transport means can be mechanical or pneumatic and convey the fiber flakes into so-called mixing boxes, in which the delivered fibers are filled in as a flake mixture.

From these mixing boxes, the fiber flake mixture is transferred to a collective transport at different speeds in order to obtain a doubling effect in order to strive for homogenization of the fiber flake mixture.

Such homogenizers are shown and described, for example, in German Patent Nos. 196 121 and 31 51 063.

The disadvantage of the first-mentioned removal and mixing process, however, is that the mixture, due to the stationary rows of bales, is unchangeable until such a row has been removed, so that the mixing process Ratio remains the same throughout this time, while the second removal and mixing process also has the inaccuracy of the amount withdrawn.

It is therefore the task to produce precise and homogeneous fiber mixtures, which can also be changed quickly as required.

It has already been proposed by the present applicant (CH patent application 03335 / 88-8) to solve this problem by forming fiber blend components with predetermined different fiber properties, which are mixed together with controllable variable component proportions to form a component blend, and in that this mixture of components depending on predetermined, respectively. ascertained, changed properties of a subsequent intermediate, e.g. a card sliver or an end product, e.g. of a yarn, respectively. is corrected.

This measure allows fibers with different fiber properties, which are determined in advance by taking samples from the fiber bale, to be mixed precisely in order to achieve the desired properties of an intermediate product, e.g. a card sliver, or an end product, e.g. to get a yarn.

There is also the possibility, e.g. by measuring fiber properties on the card sliver or yarn to determine deviations which immediately make it possible to correct the mixture in order to maintain the required properties of the card sliver or yarn.

As already indicated, the fibers have individual fiber bales of different origins of different fiber properties create. The most important fiber properties are, for example, the thickness of the individual fibers (called micronaire), the so-called stack (length of the fibers over the range from the shortest to the longest fiber, taking into account the percentage of the individual fiber lengths), the color in the sense of the basic color of the fibers (Yellow tinge), the color based on the contamination of the fiber, the fiber strength (of the individual fibers) and the extensibility of the fibers.

Depending on the intended use of the finished yarn, the fiber properties mentioned play a different role, so that the contributions of the individual components to the properties of the mixture or the yarn produced therefrom must be taken into account when mixing the fiber bales.

For example, for very fine yarns, e.g. used for women's outerwear or men's shirts, make sure that the pile is as long as possible, the fineness of the fiber is high (micronaire) and the fiber has a high strength. Other important parameters are the colors of the individual fiber provenances, which determine the appearance of the yarn. The elasticity of the individual fiber provenances also plays an important role because it influences the subsequent weaving process. In contrast, in the case of yarns for processing into denim fabrics, the stack length plays a much smaller role than in the case of fine yarns and it is essential for these yarns that the dust is removed completely, since otherwise the rotor grooves can become dirty.

If a deviation in the properties of the card sliver or yarn obtained is now determined for a specific fiber mixture with several components, this can in certain circumstances be caused by various different ones Changes in the fiber mixture are corrected. This fact leads to difficulties both in the creation of the control program and in the fact that various possible changes may not be in the spirit of spinning. For example, an automatic regulation could lead to the consumption of a certain component being too high, for which the stocks of this component are not sufficient at all. As a further example, it would also be conceivable that the regulation causes the increased supply of a certain relatively expensive component, although the same effect could also be achieved by supplying a cheaper component.

Finally, it can be seen from these examples that the task of producing precise and homogeneous fiber mixtures, which can also be changed quickly as required, should also be understood in such a way that the guidance of the spinning mill should be enabled in the simplest possible manner and Way to determine the fiber mixtures to be produced and, if necessary, to set priorities in the control process.

• I) wenigstens folgende Angaben in einer Regelung vorgege­ben werden:
  • a) die zunächst grob geschätzte erwünschte quantita­tive Komponentenaufteilung,
  • b) die Eigenschaften der Fasern der einzelnen Kompo­nenten, und
  • c) die gewünschten Eigenschaften des aus der Fas mi­schung hergestellten Kardenbandes bzw. Garnes;
• II) aus diesen Vorgaben die Regelung, entsprechend einem vorgegebenen Regelalgorithmus eine Komponentenauftei­lung errechnet, die der vorgegebenen Komponentenauftei­lung nahekommt, und die Kardenband- bzw. Garneigenschaf­ten erfüllt,
• III) die Regelung den Betrieb eines die einzelnen Komponen­ten mischenden Mischers so ansteuert, daß die errechne­te Komponentenaufteilung in dem vom Mischer gelieferten Fasergemisch erhalten wird.

To solve this specific task, it is proposed, based on the method mentioned at the outset, that

• I) at least the following information is specified in a regulation:
  • a) the initially roughly estimated desired quantitative component distribution,
  • b) the properties of the fibers of the individual components, and
  • c) the desired properties of the card sliver or yarn produced from the Fas mixture;
• II) from these specifications, the control system calculates a component distribution according to a predetermined control algorithm that comes close to the predetermined component distribution and fulfills the card sliver or yarn properties,
• III) the control controls the operation of a mixer mixing the individual components in such a way that the calculated component division is obtained in the fiber mixture supplied by the mixer.

This process enables the spinning mill management to choose the quantitative component distribution according to the inventory and according to customer requirements, taking into account both the properties of the fibers of the individual components and the desired properties of the product made from the fiber mixture. The properties of the fibers of the individual components can be determined by laboratory tests of the individual fiber balls or on-line. It is also possible to provide each fiber bale with a code that indicates the properties of the material contained therein.

These specifications simplify the control algorithm on the one hand, and on the other hand it is possible to choose the control algorithm in such a way that a concrete mathematical solution for the component distribution can be reliably found that comes close to the specified component distribution and thus fulfills the wishes of the spinning mill management.

The method just described can also be carried out by additionally specifying d) at least one control priority in the case of feature I), in the sense that compliance with at least one component component or a card sliver or yarn property has priority. With this method, the spinning mill management can ensure, for example, that the yarn produced contains at least a certain percentage of an inexpensive fiber component or has a dirt content that does not exceed predetermined limit values.

The method can then easily be carried out in such a way that a weighting is also specified for each control priority listed. This weighting can also be specified by the order of the information.

When carrying out the method, at least some of the desired card sliver or yarn properties can be measured during the card sliver or yarn production and communicated to the control system, the control system recalculating the component distribution in the event of deviations from the specification in relation to the measured properties. As a result, fluctuations in the properties of the fibers of the individual components are taken into account. For example, it can easily happen that the samples taken from the particular fiber bales are not representative of the properties of the entire bale. Such occurrences are also taken into account by the procedure according to the invention.

You can also measure additional card sliver or yarn properties in the laboratory and, in the case of disturbing deviations, also enter these into the control system, which also means these deviations when recalculating the Component distribution are taken into account.

In order to eliminate undesirable fluctuations in the control process, the properties measured during the production of the card sliver or yarn should only be taken into account by the control after the corresponding averaging. The calculation of the component division is preferably carried out according to the principle of the minimum deviations or the minimal weighted deviations from the target specification.

One way of realizing this is to carry out the calculation of the component division according to the principle of the minimum quadratic deviations or the minimum weighted quadratic deviations from the target specification. A further possibility for the calculation of the component division according to the following equation or the following control algorithm runs by the quality criterion
J (u) = _o ∫ ^∞ { x ^T (t) Q x (t) + u ^T (t) R u (t)} dt
is minimized
in which
x (t) indicates the system deviations in the form of a vector, ie the deviations of the measured properties from the desired properties,
x ^T (t) is the transform of x (t),
u (t) is the control vector that indicates the desired component split,
u ^T (t) is the transform of u (t),

Q and R are matrices with which the individual components are weighted in x (t) and u (t).

The control can also be used at the same time for setting a coarse cleaning unit which is switched between a bale removal machine and the mixer, the setting of the coarse cleaning unit influencing the card sliver or yarn properties and thereby also the calculation of the component distribution.

A coarse cleaning unit or a fine cleaning unit can also lead to a falsification of the mixture, which can only be taken into account if the regulation takes into account the effect of the cleaning unit.

For example, in the case of a contaminated starting component, it may be necessary to carry out a very intensive coarse cleaning, with relatively much of the short-staple fibers being excreted, so that the stack of the end product is rather too long with regard to the properties required. Under these circumstances, it would make sense for cost reasons to increase the proportion of a relatively short-stack, cheaper component in the mixture.

In order to take such circumstances into account, the invention therefore provides that the control system takes into account the setting of the existing cleaning unit or cleaning units when calculating the component distribution.

If, for example, intensive fine cleaning is carried out, stack reductions are to be expected, so that a change in the component mixture is appropriate in order to obtain the desired stack for the card sliver.

Even if the control does not influence the setting of the fine cleaning unit, it should at least receive information about the setting of the fine cleaning unit in order to carry out the calculation of the component distribution in a manner that is appropriate to the process.

Another advantage of the method according to the invention is noticeable when changing the range. Here, the method according to the invention provides that the control system coordinates the readjustment of the component distribution and the can change at the card exit so that the transition from one assortment to the next takes place without any significant interruption and with minimal product loss.

For example, when initiating a change of assortment, a change of can can be carried out immediately upon initiation of this assortment change or shortly afterwards at the card exit, at a time when one can be sure that the card sliver produced still has the desired properties of the previous assortment. Now a jug is used, which takes up the card sliver until the card sliver with the desired properties of the new range is obtained at the card exit. As soon as this has occurred, the regulation causes another can change, the new can taking up the card sliver of the new range.

The card sliver produced during the range change can be reused as a mixing component, ie it can be fed back into the mixer. If this takes place in smaller percentages, it does not lead to any noteworthy falsification of the desired product, and the regulation is able to keep the properties of the product within the selected provenances.

• I wenigstens folgende Angaben in einen Rechner eingegeben werden:
  • a) die Eigenschaften der Fasern der einzelnen Komponenten, und
  • b) die gewünschten Eigenschaften des aus der Faser­mischung hergestellten Kardenbandes bzw. Garnes;
• II aus diesen Vorgaben der Rechner entsprechend einem vorgegebenen Rechenalgorithmus eine Komponentenauf­teilung errechnet, die bei minimierter Abweichung von dem gewünschten Kardenband bzw. den Garneigenschaften diese wenigstens annähernd erfüllt, und ggf. eine Korrektur der errechneten Komponentenaufteilung unter Berücksichtigung etwaiger, ebenfalls in den Rechner eingegebenen Randbedingungen bzw. Sonderwünsche vornimmt und eine korrigierte Komponentenaufteilung errechnet,
• III die vom Rechner ermittelte Komponentenaufteilung bzw. korrigierte Komponentenaufteilung für die Einstellung bzw. Regelung der Zuspeisung der einzelnen Komponenten am Mischer ausgenutzt wird, um die errechnete und ggf. korrigierte Komponentenaufteilung in dem vom Mischer gelieferten Fasergemisch zu erhalten.

Another solution according to the invention of the specific task in a method of the type mentioned at the outset is characterized

• I at least the following information is entered into a computer:
  • a) the properties of the fibers of the individual components, and
  • b) the desired properties of the card sliver or yarn produced from the fiber mixture;
• II a computation of components is calculated from these specifications of the computer in accordance with a predefined calculation algorithm, which at least approximately fulfills these with a minimal deviation from the desired card sliver or the yarn properties, and, if necessary, a correction of the computed component distribution taking into account any boundary conditions that are also entered into the computer or Makes special requests and calculates a correct component allocation,
• III the component division or corrected component division determined by the computer is used for setting or regulating the feeding of the individual components on the mixer in order to obtain the calculated and possibly corrected component division in the fiber mixture supplied by the mixer.

This solution according to the invention is of particular importance because it takes into account the purchase prices of the fiber components and produces a fiber mixture whose price is at a minimum. Particular variants of this method can be found in the further claims 16 to 20. These variants of the process enable the spinning mill to run through the effects of various wishes or ideas and show in a transparent manner whether the implementation of these wishes is associated with particular disadvantages, for example too far from optimal production.

Claim 20 is of particular importance since it takes into account that the real price of the material differs from the purchase price. For example, if a material costs $ 1 / kg but contains 7% dirt (dust, shell parts, etc.). Then it has a real price of $ 1 for 930 g, equal to $ 1.075 / kg. A second component with a purchase price of $ 1.05 / kg, but which only contains 2 g of dirt, has a real price of $ 1.071 / kg and is actually less expensive than the first component, although after a direct comparison of the purchase prices another impression arises. Since the program according to the invention for calculating the ideal mixtures contains the price of the individual components as a basic specification and even minimizes the total price of the fiber mixture as a step in the optimization process, it is preferred not to use the direct purchase prices for the materials, but rather to use adjusted values which correspond to the Take into account the dirt content of the material. If the dirt is considered as a fiber property, ie as an input variable, since it ultimately also represents a property of the respective component, then the The computer determines the adjusted prices before it calculates the optimized component distribution.

The invention also encompasses devices for carrying out the methods described and claimed above, in particular using a computer which carries out the regulation.

• Figur 1 bis 5
  je eine schematische Darstellung eines erfindungsgemäßen Mischverfahrens,
• Figur 6 und 7
  eine Variante der Ausführungsart des Mischverfahrens von Fig. 5,
• Figur 8
  eine schematische Darstellung je einer Erweiterung der erfindungsgemäßen Verfahre nach Fig. 1 bis 7,
• Figur 9
  eine schematische Darstellung einer Variante des erweiterten erfindungsgemäßen Mischverfahrens von Fig. 1 bis 8, beispielsweise mit einer in Fig. 3 dargestellten Faserabtrennung,
• Figur 10
  eine Variante des Verfahrens von Fig. 2,
• Figur 11
  ein schematisches Diagramm, das zur näheren Erläuterung des Regelverfahrens dient und zwar bei einer Ausführungsform entsprechend der Fig. 2, wobei steuerbare Dosierapparate am Ausgang jeder Komponentenzelle vorgesehen sind, wobei das gleiche Regelverfahren auch für die Ausführungsformen nach Fig. 3 und 5 sowie mit gewissen Abwandlungen für die andere Ausführungsform verwendbar ist,
• Figur 12
  ein weiteres schematisches Diagramm ähnlich der Fig. 11, jedoch für das Regelverfahren mit einer Ausführung ähnlich der Fig. 9 mit einer zwischen der Ballenabtragmaschine und dem Mischer angeordneten Grobreinigungseinheit, jedoch mit der weiteren Besonderheit, daß zwei Feinreinigungseinheiten vorgesehen sind, die aber im Gegensatz zu Fig. 9 nach dem Mischer angeordnet sind,
• Figur 13
  ein schematisches Diagramm, das dem der Fig. 12 ähnlich ist, jedoch noch genauer auf die Ausführung der Fig. 9 gerichtet ist, wobei zwei Feinreinigungseinheiten vorgesehen und direkt nach der Grobreinigungseinheit angeordnet sind, und
• Figur 14
  ein noch weiteres schematisches Diagramm, ähnlich der Fig. 11 bei dem aber die Regelung auch ausgelegt ist, den Kannenwechsel am Ausgang der Karden mit einem Sortimentswechsel zu koordinieren.

The invention will be explained in more detail with reference to drawings which only show execution methods, with FIGS. 1 to 10 showing different possibilities for the construction of a plant for the removal and mixing of fibers of different provenances and reflecting the disclosure from the earlier Swiss patent application CH 03335 / 88-8 11 to 14 and the tables in FIGS. 16 to 20 show different variants of the control method according to the invention. More precisely show:

• Figure 1 to 5
  each a schematic representation of a mixing process according to the invention,
• Figures 6 and 7
  5 shows a variant of the embodiment of the mixing process from FIG. 5,
• Figure 8
  1 shows a schematic representation of an extension of the method according to FIGS. 1 to 7,
• Figure 9
  2 shows a schematic representation of a variant of the extended mixing method according to the invention from FIGS. 1 to 8, for example with a fiber separation shown in FIG. 3,
• Figure 10
  a variant of the method of Fig. 2,
• Figure 11
  a schematic diagram that serves to explain the control method in more detail in an embodiment corresponding to FIG. 2, wherein controllable metering devices are provided at the outlet of each component cell, the same control method also for the embodiments according to FIGS. 3 and 5 and with certain modifications can be used for the other embodiment,
• Figure 12
  a further schematic diagram similar to FIG. 11, but for the control method with an embodiment similar to FIG. 9 with a coarse cleaning unit arranged between the bale removal machine and the mixer, but with the further special feature that two fine cleaning units are provided, but in contrast to 9 are arranged after the mixer,
• Figure 13
  12 shows a schematic diagram, which is similar to that of FIG. 12, but is directed even more precisely to the embodiment of FIG. 9, two fine cleaning units being provided and arranged directly after the coarse cleaning unit, and
• Figure 14
  yet another schematic diagram, similar to FIG. 11, but in which the control is also designed to coordinate the change of cans at the exit of the cards with a change of product range.

FIG. 1 shows a number of conveyor belts 1 for receiving fiber bales 2 which are removed by fiber bale removal elements 3.

The respective fiber bale removal member moves on stationary rails which are arranged, for example, in the diagonal direction of the fiber bales 2 located on the conveyor belt. Such a device, identified here by the reference number 20, is known in principle from the applicant's Swiss Patent Specification No. 503809. As a variant to this, the device shown and described in the applicant's Swiss patent application with the number 00399 / 88-8 could be used, in which the removal member 3 is attached to a removal device (not shown) that can be moved back and forth along horizontal bales 2 along the bale 2, can be moved up and down, as well as tilted for diagonal removal.

The removal rate in both removal devices can be controlled by changing the displacement speed of the fiber bale removal member 3 along the above-mentioned diagonal path, as well as by changing the feed speed of the fiber bale 2 by means of a variable speed of the individual conveyor belt 1.

The fiber flakes detached from the removal drum 4 are pneumatically in a manner known per se cal delivery line 5, which is not described here, transported away.

With the help of this pneumatic conveying line 5, the fiber flakes are conveyed into a mixer 6 and mixed therein to form a uniform mixture.

The quantities conveyed into the mixer 6 by means of these individual pneumatic conveying lines 5 are referred to hereinafter as fiber flake components or simply components.

Batch mixers or continuous mixers can be used as mixers; depending on the quantities mentioned are individual weight batches (kg) or a running quantity per unit of time (kg / h).

For the sake of simplicity, the delivery lines 5 in FIG. 1 schematically open directly into the mixer 6, which is also shown schematically, but in practice this can vary depending on the type of mixer. For example, air-fiber separators can be used to separate the respective fiber-air mixture from one another, so that the fiber flakes can fall into the mixer in free fall, while the air can be led into an exhaust air line. Such separators are well known from practice and are therefore not shown here in particular.

The stated quantities of the aforementioned individual fiber flake components which are added to the mixer 6 are controlled by a controller 7 on the basis of a control program.

Such a control program can be a computer program be, which has a component mixing program that adapt to adapt to changes in the mixture, respectively. is changeable.

Another variant would be a digital control for each component, in which the performance of the individual components was selected manually. could be changed.

The functions relevant to the removal performance of the components, e.g. the feed speed of the respective conveyor belt 1 or the removal movement of the fiber bale removal member 3 is controlled by one or the other controller.

It goes without saying that the pneumatic conveying lines do not have to convey the removed product directly into the mixer, but that mechanical conveying elements are interposed, for example conveyor belts. In such a case, the fiber air separators mentioned place their fiber product in such mechanical conveying elements.

Each fiber removal member 3 is connected to the controller 7 via a control line 8 and each conveyor belt 1 via a control line 19.

The three incoming control lines in the controller 7 will be described later.

Figure 2 shows a variant of Figure 1, but in which the same elements have the same reference numerals. In it, the pneumatic conveyor lines 5 promote the removed fibers. Fiber flakes, also called product, are not directly in the mixer 6 but in component cells 9, from which the product filled therein is in each case discharged by means of a discharge device 10 and is fed into the mixer 6 by means of a subsequent metering device 11.

Depending on the type of discharge device 10, this can also take on the dosing function as a variant.

The discharge rate from the individual component cells 9 is controlled by a controller 7.1, which controls the individual metering devices 11 and 11 by means of control lines 12. as a variant that controls the discharge apparatus 10.

In the first-mentioned disposition, the metering devices 11 can each be controlled by means of a control line 13 via the dispensing devices 10 in order to coordinate the dispensing with the metering. The discharge devices could also be controlled directly by the controller 7.1.

The component cells 9, which can be designed according to the German patent application P 39 13 997 2, are filled by the elements 1 to 5 already mentioned for FIG. 1, the use of two fiber bale rows, each with the elements 1 to 4, being chosen only as an example is. In practice, several rows of fiber bales or just a single row per component cell 9 could also be selected. Such a decision depends on the number or mix of provenances per row of bales, which are to form a mixed component to be placed in a corresponding cell 9.

Furthermore, the filling up of the component cells 9 is, for example, by full provided in each cell level indicator 14 and controlled by vacancy indicator 15 by means of a controller 16. For this purpose, the control 16 for the reciprocating movement of the removal elements 3 is connected by control lines 17 each to the fiber bale removal elements 3 and by control lines 18 each to the drive motors of the conveyor belts 1.

FIG. 3 shows a further embodiment in which the same elements already shown and described with FIG. 2 have the same reference numerals. This applies to the fiber bales 2, the component cells 9, the discharge apparatus 10, the metering apparatus 11, the mixer 6 as well as the control 7.1 and the control lines 12 and 13.

For the removal of the fiber bales 2, which are here directly on the floor, these are also set up in groups which correspond to the respective provenance of the fiber bales. The removal is done by a mobile fiber bale removal device 20, which runs along the fiber bale groups and from the surface of fibers or. Removes fiber flakes. Such a device is known in the spinning industry under the name "Unifloc" and is sold by the applicant worldwide.

As is known, this fiber bale removal device 20 conveys the removed fibers via a pneumatic conveying line 21 into the corresponding component cells 9.

As already described for FIG. 2, the component cells 9, fullness detectors 14 and vacancy detectors 15, which input their signals to a controller 22. This control is via a control line 24 connected to the fiber bale removal device 20 and controls the removal of the fiber flakes from the corresponding fiber bale groups for the filling of the corresponding component cells 9.

As shown schematically in FIG. 3, the fiber bale removal device 20 has a fiber removal member 23 which is known per se from the Unifloc and which removes the fibers from the bale surfaces by means of a removal drum (not shown) rotating therein.

It is also known that the fiber bale removal member 22 can be rotated through 180 degrees as indicated by the arrows M that the fiber bale removal member can remove the fiber bale group 2 on the opposite side. This enables either one of the opposite fiber bale groups to be used as a reserve fiber bale group, or that with an automatic rotation of the fiber bale removal device 20 mentioned above, both opposite bale rows can be removed with a predetermined variation.

FIG. 4 shows a variant of FIG. 3, so that the elements already described and shown with FIG. 3 have the same reference numerals.

The difference between that shown in FIG. 3 and FIG. 4 is that not only a single fiber bale removal device 20 is provided, but one for each of the four fiber bale groups lying opposite one another in two rows.

Accordingly, the control is with 22.1 instead of with 22 characterized, since four individual fiber bale removal devices 20 are each to be controlled separately by means of the corresponding control line 24. Likewise, a pneumatic conveying line is provided for each fiber bale removal device 20, which is accordingly identified with 21.1 instead of 21 and each opens into a component cell 9.

FIG. 5 shows an arrangement similar to FIG. 1, in which instead of the single conveyor belt 1 per bale group of FIG. 1, a conveyor belt 30 with a purely conveying function and a conveyor belt 31 with a conveying / weighing function is provided per fiber bale group.

The weighing function of the latter conveyor belt can be provided, for example, by supporting the axes of the deflecting rollers of the conveyor belt 31 on pressure sockets 32 known per se, each of which emits a signal 33 corresponding to the weight, which signal is sent via a control line 33 to a controller 7.2 processing the signals is forwarded. The processing of the aforementioned signals consists in that the controller 7.2 uses these to generate the control signals which control the motors of the conveyor belts 30 and 31 mentioned and the removal elements 3 via control lines 34.

Of course, other weighing systems can also be used, which can be combined with conveyor belts.

Furthermore, the elements already described and shown for FIG. 1 have the same reference numerals Mistake.

In operation, the controller 7.2 controls the fiber removal elements 3 and the conveyor belts 30 and 31 at predetermined speeds in order to remove fibers from the fiber bales 2, which are conveyed into the mixer 6 by means of pneumatic conveyor lines 5.

Each fiber bale removal member 3 of the individual fiber bale groups conveys a predetermined amount, controlled by the control 7.2, into the mixer 6. This predetermined amount to be removed (bp / h) per bale group is monitored and monitored by the respective weighing conveyor belt 31 or by the pressure dose weighing device 31 converted into signals and delivered to the controller via the control lines 33. If the quantity (kg / h) removed per fiber bale group does not match the specified quantity, the control adjusts the quantity to be removed until it matches the specified quantity.

Measurements are always made via the measuring device 32 when the fiber bale removal member is at a standstill for a brief moment at the turning point of the back and forth removal path.

In this type of removal, the fiber bale removal member 3 always moves back and forth on the same path, essentially lying in the diagonal of the fiber bale to be removed. back and forth. The amount (kg / h) of the fibers to be removed from the bales is generated by means of the feed speed of the conveyor belts 30 and 31 and removal member 3.

The controller 7.2 can be an electronic controller be based on the analog technology or a microprocessor, by means of which the different quantities removed per bale group can be set and adapted by the signals of the control lines 33 and input signals explained later.

FIGS. 6 and 7 show a weighing system similar to FIG. 5, FIG. 7 being a plan view of FIG. 6, in the direction of the arrow A.

From Figure 7 it can be seen that this is a number of rows of bales, respectively. Bale groups that are arranged side by side and each form a mixing component. As shown in FIG. 6, the fiber bales 2 each lie on a conveyor belt 40 and a weighing conveyor 41 connected to it. Each weighing conveyor 41 can be supported, analogously to the weighing conveyor 31 in FIG. 5, on pressure measuring cells 42, of which a signal corresponding to the weight is used a control line 43 is delivered to a controller 44.

The fiber bales 2 located on the weighing conveyor belt 41 are removed by fiber bale removal device 48 in accordance with Swiss Patent Application No. 00399 / 88-8, which has already been mentioned in connection with FIG. The difference essentially consists in a long fiber bale removal member 49, which extends over the predetermined number of bale rows, with a removal drum 51 which simultaneously removes fibers from all of the predetermined bale rows shown in FIG. 7.

Another difference between this removal method and that described for FIG. 1 is that that the fiber removal member 49 removes in an oblique removal path, which essentially corresponds to the diagonal of a predetermined number of fiber bales 2 lined up, for example, as shown in FIGS. 6 and 7, of four fiber bales 2.

However, it goes without saying that a different number of bales can also be removed obliquely in this way, for example only one, as is shown with FIGS. 1 and 2.

It also depends on the possible length of the removal member 49, how many fiber bales can be lined up next to one another in order to be removed at the same time.

The fiber material removed by the fiber removal member 49 is conveyed in a pneumatic conveying line 50 which, according to the invention, opens into a continuous mixer 45. As described for FIG. 1, the conveying line 50 can end in a separator (not shown) which discharges the product into the mixer 45 .

Furthermore, the fiber bale removal device 48 is controlled by the controller 44 via the control line 46 with respect to the driving speed.

Another control line 47 is used to control the drive motors of the deflection rollers of the control belts 40 and 41.

It goes without saying that the deflecting rollers of the conveyor belts 40 and 41 (not particularly marked) have a separate drive motor for each bale group sen, ie that each motor has a control line 47 to the controller 44 separately.

In operation, the controller 44 controls the back and forth movement of the fiber bale removal device 48 along the bales located on the weighing conveyor 41 and the up and down movement of the fiber bale removal member 49 on the device 48 during the aforementioned back and forth movement, so that the fiber bale as shown in FIG. 6 be removed in an inclined direction substantially corresponding to the diagonal of the four bales 2.

This removal movement always runs in the same path and at a predetermined speed, so that the removal amounts (kg / h) of the individual fiber bale groups can be selected differently by the individual feed speeds of the conveyor belts 40 and 41. These different feed speeds of the individual bale groups correspond to a removal program with different amounts (kg / h) of the individual bale groups to be removed in order to obtain the mixture mentioned.

The drive motors for the conveyor belts 40 and 41 are advantageously drum motors which are installed in the deflection rollers of the conveyor belts. Such drum motors can be operated at different frequencies by means of frequency inverters, i.e. are driven at different speeds, which is part of the controller 44.

Likewise, as in all cases in this application and particularly mentioned for FIG. 5, the controller 44 can be an analog or digital controller by means of wel the quantities of the individual components are controlled. These quantities are corrected by means of the pressure measuring cell signals, which are input through the control line 43 of the control 44, if the individual component quantity does not correspond to the target specification.

FIG. 8 shows an extension of the previously described method, in which it is shown that after the mixer 6, the product coming from this mixer is put into a so-called blowroom 60, in which cleaning machines known per se are used.

The blowroom 60 can contain so-called coarse cleaning machines 61 and fine cleaning machines 62. This blowroom, like the previous one, is only shown schematically.

The same applies to the card 63 following the blowroom, which can be a card known per se, for example the card C4 sold worldwide by the applicant.

This card 63 is provided with a control 64, which is known per se and controls the card functions, which, among other functions, also has the function of ensuring the uniformity and the amount (kg / h) of the card sliver.

After the card, seen in the direction of belt feed, in front of the card sliver deposition (not shown), the card sliver is checked by a color sensor 65 and by a sensor for measuring the fiber fineness 66.

It should be mentioned in advance that either both sensors or only one or the other can be used.

In the case shown in FIG. 8, the color testing device 65 outputs a signal 67 corresponding to the color of the card sliver and the fiber fineness testing device 66 outputs a signal 68 corresponding to the fiber fineness to the control devices 7; 7.1; 7.2; 44 mentioned in connection with FIGS. 1 to 7 , which each control the control of the individual fiber components. A further signal 81 corresponding to the card sliver quantity (kg / h) is also sent from the card control 64 to the controls 7; 7.1; 7.2; 44 entered. These three signals are compared by the aforementioned controls with the setpoint for the sliver color entered in the control, the setpoint for the fiber fineness and the setpoint for the power, so that if there are any deviations during operation, these deviations by changing the component mixture and performance can be remedied.

The product discharged from the mixer 6 is conveyed to the blowroom 60 via a conveyor system 69 and to the card 63 from the blowroom 60 via a conveyor system 70. Such conveyor systems can be mechanical or pneumatic, it is also known per se that conveyor systems exist between fine cleaning and coarse cleaning machines.

The method according to the invention is likewise not restricted to a single blowroom 60 and a single card 63 after the mixer 6, but a plurality of blowrooms 60 can either be after the mixer 6 and several cards 63 are loaded with the product of the mixer 6 or if a blowroom is provided after the mixer 6, several cards 63 can be loaded with the product of the blowroom 60.

If several cards are planned. A color test device 65 and / or a fiber fineness test device 66 can optionally be provided after each card, or there is also the possibility if several cards process the same product, that only one so-called guide card has these latter two test devices.

FIG. 9 shows the possibility of providing the blowroom 60 between the fiber removal and the component cells 9, so that an already cleaned fiber material in the component cells 9 is available for the mixture.

The conveying device from the fiber bale removal device 20 to the blowroom 60 basically corresponds to the pneumatic conveying line 21, whereby in this case too pneumatic conveyance is not mandatory but can be mechanical.

The conveyor between the blowroom 60 and the component cells 9 can also be a pneumatic conveyor line, as indicated at 21, but it can be any conveyor system. The method according to the invention is not restricted to any conveyor system.

Likewise, the provision of the blowroom 60 is not restricted to the combination with the arrangement from FIG. 3. It is understood that everybody's fiber components arrangements shown in the figures, with the exception of FIGS. 6 and 7, can first be cleaned and then get into the mixer 6. It is only a matter of effort, since a cleaning shop must be provided for each of the components in FIGS. 1, 2, 4 and 5.

FIG. 10 shows a variant of the method of FIG. 9, in that the blow room is divided into a rough cleaning with the cleaning machines 61 and one into a fine cleaning with the fine cleaning machines 71, each of which is preceded by a storage container 72 (for the sake of simplicity only one is marked).

The fine cleaning machines 71 are started or stopped by a controller 73, namely stopped on the basis of a vacancy indicator 74 and started on the basis of a full status indicator 75 (only one identified in each case). These full and vacancy detectors emit their signals to the control 73 via the lines 76 and 77.

The coarse cleaning machines 61 are fed by means of a fiber transport 78, which can correspond to the pneumatic conveying line 21 from FIG. 9 or to any fiber conveying known per se.

The same applies to the fiber transport 79 between the coarse cleaning machine 61 and the storage containers 72.

The fine cleaning machines each pass their products on to a component mixing cell 9, as has already been described for FIGS. 2 to 4 and for FIG. 9.

Accordingly, the other elements already described are identified by the same reference numerals and are not further described for this figure.

In operation, the components are cleaned individually, accordingly, the vacancy detectors 15 of the individual component cells 9 request the removal of fibers from the corresponding fiber bale group a or b or c or d in order to clean these removed fibers in the coarse cleaning machine and pass them on to the corresponding storage container 72 , which delivers the specified component to the subsequent fine cleaning machines 71.

This product request by the empty detector 15 occurs because the corresponding fine cleaning machine no longer supplied a product, since the empty space detector 74 in the storage container 72 had also reported empty space. Correspondingly, the corresponding group a to d is removed until the corresponding fullness indicator 75 reports fullness to the removed component. The corresponding fine cleaning machine can thus be put into operation again until the fullness detector 14 reports fullness again to the corresponding component cell 9.

The fiber transport 80 between the mixer 6 and the card 63 can correspond to a fiber transport which is identified and described as 70 in FIG.

It also applies to this variant that a mixer 6 can operate several cards, so that the Fiber transport 80 transports the product delivered by the mixer to the appropriate number of cards.

The control method will now be explained in more detail, first with reference to FIG. 11, which is particularly designed for the embodiment according to FIG. 2. In order to explain the correspondence between FIGS. 11 and 2 in more detail, the same reference numerals have been used for the same parts. From Fig. 11 it can be seen that the fiber bale removal device 20 removes various components and delivers them to the respectively assigned component cell 9 of a mixer 6. In contrast to the four components of the embodiment in FIG. 2, eight different components are provided here, but the principle is the same. The metering devices 11 of the individual component cells 9 are not shown in FIG. 11, but are controlled by the control 7.1 via control lines 12 in accordance with the embodiment according to FIG. 2. The mixed product of the HF mixer then leads to a coarse cleaning unit 61 and the coarsely cleaned product is then led to a first fine cleaning unit 62.1 and then to a further fine cleaning unit 62.2. These cleaning units are not shown in the embodiment of FIG. 2, but they can also be provided there in the same way. The finely cleaned starting product of the fine cleaning unit 62.2 is then fed into the filling shafts of six cards 63.1 working in parallel.

Two of the six cards are provided with fiber fineness measuring devices (Micronaire), the output signals 68 of which lead to the control 7.1. Two further cards are provided with color checking devices 65 for the on-line measurement of the color of the card sliver, the corresponding signals 67 also being fed into the regulation 7.1. Furthermore, a further signal 81 corresponding to the card sliver production (kg / h) is fed into the control by the card control 7.1 fed.

Other parameters measured on-line can also be taken into account by the regulation 7.1, for example measurements of the stack or the stretchability of the ink ribbon or else dirt content, fiber strength etc.

The control 7.1 consists of two main blocks (100, 101), the block 100 recording the input of the spinning mill management, for example on an input keyboard (102), and calculating the actual control parameters from this. Provenance data for the individual fiber components in the individual shafts 9 of the mixer are initially described in more detail on the keyboard 102. These components are marked in FIG. 11 with X1 to X8 and for each component the control 7.1 receives data, for example about the fineness of the fibers (micronaire), the stack of fibers, the degree of contamination, the strength, etc. These details are in the memory represented by field 104. The arrow 106 indicates that the corresponding data can not only be entered manually, but possibly via a line from the bale management, which is shown here as field 108. Here, for example, the field 108 could be a code reading device that reads coded information about the properties of the fibers of the respective bales of the individual provenances and feeds the corresponding signals via line 106 into the regulation 7.1.

In addition to this information, control 7.1 receives a desired idea of the spinning tour via the component division of the individual components X1 to X8 via the input keyboard 102. This desired idea of the division of components is recorded in a memory, which is identified by 110.

When compiling the desired component distribution, the spinning mill management can take into account, for example, the inventory of the individual components and the need to recycle a certain amount of waste. In the example shown, the exit is specified as component X8, of which a 3% share should appear in the card sliver according to your wishes. Of course, the desired distribution of components must reflect the inventory on the one hand, but also the desired card sliver product on the other.

Regulation 7.1 also receives information on the desired card sliver properties, i.e. to the permissible ranges of these card sliver properties, which are stored in a memory, which is represented by reference number 112. The desired card sliver properties can be, for example, properties such as fineness, stacking, color, stretchability, price, etc., the number of properties is not limited, but the controller algorithm only has to be designed in such a way that all the properties entered can also be taken into account.

Field 114 represents a priority memory that contains a specific order of control priorities. In the example shown, the fineness of the card sliver is in the first place, in the second the stack, in the third place the need to use 3% waste in the form of component X8, in the fourth place the color and in the fifth place the wish if possible 25% Component X1 to process, because this component has been purchased inexpensively. In the example shown, the order of the information also represents a weighting of the rule priorities. However, there is also the possibility for each priority specify a special weighting. Properties that are not specifically listed as priorities are then weighted by the control with a priority of zero.

The content of the respective memory fields 104, 110, 112, 114 can preferably also be displayed on a screen, so that the user can immediately recognize which information is currently decisive for the regulation. If desired, all fields can be displayed on the screen at the same time or selectively only individual fields, possibly with additional comments, if this is desired by the user.

The control 7.1, or more precisely the microprocessor 100, then calculates a component distribution which, taking into account the provenance data of the individual components and the control priorities, possibly taking into account the weighting of the control priorities, provides a card sliver with properties lying within the desired ranges and the desired component distribution comes closest. The calculation of this component division is indicated by the field 116 of the microprocessor 100. These control parameters, ie the division of components, which is preferably expressed in mass flows, are calculated in such a way that the sum of the deviations, weighted according to priorities, between the default values and the actual values is as small as possible. The values from the desired component distribution are also considered as default values, usually with a low priority weighting. This special method, ie the consideration of the values of the desired component division as a default value, ensures according to the invention that the control loop is always mathematically overdetermined, so that an optimization with a clear result is possible.

The controller parameters or mass flows of the individual provenances X1 to X8 calculated in field 116 then form the target values for a control circuit 118, which ensures that the corresponding mass flow values are actually maintained.

Since it is possible to measure some technological values of the card sliver on-line, for example the fineness (micronaire), the color and also the production, these card sliver properties can be included in the control 118, which can be seen with the corresponding signals 68, 67, 61 is indicated in FIG. 11. If these values lie outside the tolerance ranges specified in the memory 112, the component parts X1 to X8, i.e. the corresponding mass flows are recalculated according to the principle of minimal weighted deviations from card sliver properties and component distribution, taking into account the actual deviations from the card sliver properties, at least as far as the micronary and color values are concerned. These newly calculated corrected words X1 to X8 are then used for the mass flow control in the mixer 6. This regulation also takes into account the fact that there is a dead time TZ between the discharge of the components from the metering apparatus of the mixer 6 to the discharge of the corresponding card sliver from the cards. In the schematic diagram of FIG. 11, it is assumed that the coarse cleaning unit 61 and the fine cleaning units 62.1 and 62.2, and also the card, do not cause any stack damage. It is also assumed that the dirt is separated as completely as possible, this separation being able to take place both in the cleaning units 61, 62.1 and 62.2 and in the individual cards 69.1.

But even if the cleaning units, especially the fine cleaning units, some stack damage, ie Shorten the stack, this will be reflected in the card sliver. Since online measurement of the stack is currently relatively difficult, samples from the card sliver can be examined in the laboratory to determine the actual stack. If the actually measured stack deviates from the value calculated in field 116, this is on the one hand an indication that either the fine cleaning units or the cards have caused this stack shortening. The actually measured value for the stack and possibly other measured values as well as the micronaire and color values in the controller 118 can also be taken into account in the calculation of new components X1 to X8 based on the principle of the minimum weighted deviations from card sliver properties. This also applies to all other technological values that can be measured in the laboratory.

The coarse cleaning machine is a very gentle type of cleaning with regard to fiber damage, but essentially only eliminates the coarse impurities, so that the rather finer impurities have to be removed in the aggressive fine cleaning machines, but this includes the possibility of fiber damage. In the case of rough cleaning, there is also the possibility that relatively short-staple fibers are removed with the dirt, i.e. are lost, so that the setting of the coarse cleaning unit can also cause a change in the stack of the finished card sliver.

FIG. 12 shows one possibility for taking this into account, in which, in contrast to FIG. 11, the coarse cleaning unit 61 is arranged between the fiber bale removal device 20 and the mixer 6. The control 7.1 is essentially the same as the corresponding control of FIG. 11, only the microprocessor 100 receives a message via line 120 about the actual setting of the Coarse cleaning unit. This setting is taken into account when calculating the control parameters in field 116, with regard to the possible separation of short-staple fibers and of coarse dirt. It is also possible to control the coarse cleaning unit via line 122 from 116 so that short-staple fibers and / or dirt are eliminated.

Since the coarse cleaning unit is inserted between the fiber bale removal device 20 and the mixer 6, it may also be sensible to measure the provenance data on the basis of samples taken from the cells 9 and to first enter these values into the memory 104, since in this way the effect of the Coarse cleaning unit with regard to the stack of the individual components and with regard to the dirt content of the individual provenances can easily take into account.

In this example, the fine cleaning units 62.1 and 62.2 are switched on one after the other between the mixer 6 and the cards 63.1 operated in parallel. In this example, the sensors and actuators for the coarse cleaning unit are connected to the computer 100, but this is not absolutely necessary. A separate control can be provided for the cleaning machine, but here it is important that the product is examined after the rough cleaning unit in order to take into account the effect of the unit with regard to stack changes and dirt removal in the individual components.

13 shows that it is also possible to use the fine cleaning units 62.1 and 62.2 likewise between the fiber bale removal device 20 and the mixer 6. In this case too, the sensors and actuators can be used for the fine cleaning be connected to the computer 100. Therefore, the computer can find out the actual settings of the fine cleaning units via lines 124, 126 and therefore also take into account the effect of the fine cleaning unit with regard to dirt removal and fiber damage, and stack reductions. Via the lines 128, 130, the computer 100 can also control the fine cleaning units in such a way that the desired degree of dirt removal takes place and that the stack reductions that occur remain within predetermined limits.

13, it is also possible to provide the cleaning units 61, 62.1 and 62.2 with their own controls and to determine the effect of these units on the individual provenances by taking samples from the component cells 9 of the mixer 6. Otherwise, it is easy to see from FIG. 13 that the control 7.1 is carried out in accordance with the control of the diagrams of FIGS. 11 and 12, which is why the same reference numerals are used for the same parts.

Finally, FIG. 14 shows a control method corresponding to FIG. 11, with an additional automatic change of assortment from assortment control 7.1.

In preparation for a change of assortment, the card sliver properties, control priorities and the desired component distribution, which have been adapted to the changed yarn requirements, are first re-entered and new control parameters are calculated from them. After the "start pressure" to change the range, the following procedure then takes place.

First, the metering devices in the mixer 6 are reset for the individual components so that the new sorti appears at the outlet of the mixer. A production-dependent material throughput time is then waited, which in a practical example is approximately 2 minutes, and a can change is then automatically initiated via control line 132. This means that the cans at the exit of the cards, which are (partly) filled with the old range, are exchanged for new cans, which then take up the card sliver of a transition range with changing properties. The card's sensors, for example those for micronaire and color, can be used to determine how long these property changes last or when the properties have stabilized. The control system carries out the corresponding examination on the basis of the signals received via lines 87 and 67. As soon as it is certain that the change in properties has stabilized, an automatic can change is carried out again for all cards. The content of the cans partially filled during the change in properties is to be regarded as a strip outlet and can be used, for example, for the outlet component X8. The cans used after the stabilization of the change in properties are given a card sliver from the new range, which is then spun into yarn.

Instead of using the signals from the Micronaire and color sensors for the second automatic can change, it is also easy to wait a sufficient amount of time since the metering devices were reset for the new range before the can change is initiated. However, this tends to lead to more tape exit, since you have to work with larger safety factors.

11 to 14, the component components, ie mass flows X1 to X8, are fed via the control line 12 to the metering devices of the individual Component cells of the mixer 6 created. However, it will be clear that the corresponding signals are also used, for example, in the case of arrangements according to FIG. 1 for controlling the bale removal elements 3 and / or the conveying elements 1, so that the control can also divide the range of components in this way.

In order to allow the procedure in the inventive determination of the respectively desired component division as the basis for the subsequent setting of the bale removal and mixing system, a preferred method is now shown in more detail including the calculation steps carried out.

The starting point for this process variant is the desire to mix a predetermined number of fiber components, the quality characteristics and the price of each fiber component being known. This process also presupposes that the quality of the resulting mixture is at least known as a wish. The quality of the mixture means the properties of the fiber mixture, which are reflected, for example, in the properties of the card sliver or the finished yarn. To be more precise, the composition of the mixture should be determined that comes as close as possible to the imagination and is minimal in terms of price. The mathematical side can be explained as follows:

a) Notation agreements

- Scalars and vectors are represented symbolically with small letters, matrices with large letters.

- x = [x1, x2, .., xn]; x denotes a line vector with the vector components x1, x2, to xn.

y = [y1, y2, ..., yn] ′; y denotes a column vector with the vector components y1, y2, to yn.

It can be seen from the context whether it is a column or row vector.

- Meaning of special symbols:
′ .. transposition
* .. multiplication

b) Problem analysis

• (i) 0 = < ci = < 1 für die ganzen Zahlen i = 1 .. n
• (ii) 1 = c1 + c2 + .. + cn

It is assumed that linear mixing laws apply to all quality features. The mix quality q is calculated according to G1. (1).
Q = QK ^* c (1)
q .. is a vector, the components of which describe the individual quality characteristics of the mixture,
QK .. is a matrix, the elements of which describe the individual quality characteristics of the components to be mixed. In the illustration
QK = [q1, q2, .., qn]
qi, with integers i = 1 .. n, are the column vectors of the component qualities,
c .. is a vector, the components of which describe the individual proportions of the components to be mixed.
Two properties of c = [c1, c2, .., cn]:

• (i) 0 = <ci = <1 for the integers i = 1 .. n
• (ii) 1 = c1 + c2 + .. + cn

The mix price, calculated as a dot product according to G1. (2)
p = pK ′ * c (2)
p .. Price of the mixture, for example in francs / kilogram
pK .. is a vector whose components describe the prices of the individual components to be mixed.

Discussion of Eq. (1)

Eq. (1) is a linear equation in c. If one leaves aside the constraints that the mixture vector c has to satisfy, then linear algebra teaches that Eq. (1) has a solution only if and when q (vector of the mixture quality) as a linear combination of the vectors qi ( Component quality vectors) can be represented. Equation (1) can only be solved if: Rank (QK) = Rank (QK, q).

Is Eq. (1) solvable and if rank (QK) = n-r, this means that r components of the mixture vector c are freely selectable.

If Eq. (1) cannot be solved (e.g. if there are more quality features than mixture components), then at least the mixture c that comes closest to the required quality should be determined. The method to be used is well known, it is the equalization calculation.

Precise task

We are looking for an algorithm that delivers the mixture vector that is as close as possible to the desired mixture quality and that can be physically implemented by satisfying the constraints. Individual components of the mixture vector should be fixed. The price of the mixture should be as low as possible. The problem can generally only be solved with a certain willingness to compromise. The smears accepted with regard to individual quality characteristics of the mixture should be able to be specified with a weighting.

c) The solution

Eq. (3) defines a loss function v (c) with which the generalized losses are measured, which result from the fact that the desired quality characteristics of the mixture are not completely achieved and that a non-zero price has to be paid for the mixture is. The price is, at least from an accounting point of view, on the issue side or the loss side. The deviations of the achievable mixture quality from the desired quality and the mixture price are also weighted:
v (c) = (QK ^* - q) ′ ^* W ^* (QK ^* c - q) + w ^* c ′ ^* p ^* c (3)
W .. positive semidefinite diagonal matrix for weighting the quality deviation of the mixture from the desired quality,
P .. positive definite diagonal matrix, the elements of which are the prices of the components,
w .. scalar to weight the price influence.

All mixture vectors should also satisfy the constraint Eq. (4). This equation expresses that the sum of the mixture proportions must be 1 (see Chapter 3, property (ii)).
g (c) = 0 = k + e ′ ^* c (4)
e .. vector of the same dimension as c, the elements of which are all 1,
k .. scalar (k = -1 if all components of c are freely selectable)

We are looking for the mixture vector c for which the function value v (c) is minimal and which fulfills the constraint Eq. (4). In addition, the constraints according to Eq. (5) must also be observed:
0 = <ci = <1 for the integers i = 1 .. n (5)

The task is solved step by step.

Step 1

The system of equations consisting of equations (3) and (4) is solved according to known rules of differential calculus. This leads to an optimal mixture vector cl. If the secondary conditions (5) are not violated, the task is solved, otherwise the second step is carried out.

2nd step

A component of c, whose corresponding component of c violates Eq. (5), is manually set to a fixed value that is compatible with Eq. (5). The mixture vector c is thus restricted by one dimension. In Eq. (4) k is determined so that the meaning of that equation remains valid. Then the first step is carried out again, which leads to an optimal mixture vector c2.

Steps 1 and 2 are carried out until the secondary conditions Eq. (5) are not violated in step 1. The process terminates at the latest when all components of c are set to fixed values by hand.

The attached tables in FIGS. 15 to 20 show what the mathematical calculations discussed above based on a few concrete examples now look like.

• 1. die mittlere Stapellänge der Komponenten,
• 2. die Feinheit der Komponenten in Mikronärwerten ausgedrückt,
• 3. der Farbwert FAK von den einzelnen Komponenten,
• 4. der Farbwert FBK von den einzelnen Komponenten, und
• 5. der Preis der jeweiligen Komponenten, beispielsweise in Franken pro Kilogramm, vorzugsweise unter Anwendung von korrigierten Werten nach Berücksichtigung des Schmutzanteils, der ausgeschieden wird.

The table in FIG. 15 initially shows in the first line the time at which the calculation process was carried out, namely by specifying the year, month, day, hour, minute and second. This information is not of particular importance for the method according to the invention, it merely enables the calculation to be assigned to the sequence in operation. It is essential that in the example shown here there are six different components x1 to x6 (instead of the eight components x1 to x8 of the previous examples), which is why the upper part of the table has six columns. Five different properties are given for each component in the present example. The five properties are as follows:

• 1. the average stack length of the components,
• 2. the fineness of the components expressed in micronary values,
• 3. the color value FAK of the individual components,
• 4. the color value FBK of the individual components, and
• 5. the price of the respective components, for example in francs per kilogram, preferably using corrected values after taking into account the proportion of dirt that is excreted.

At this point it should be emphasized that this is only an example. In practice, other or further or fewer properties of the individual components can also be taken into account with the same method, even with a different number of components.

The middle part of the table shows the result of a first optimization using the mathematical method described above. The same values are also given for the mixture itself, i.e. Stack of the mixture, fineness of the mixture, color value a of the mixture, color value b of the mixture and price of the mixture. The values given first in this display are the actually calculated values (step 1). Particular attention should be paid to the S and W values given in brackets after each entry. Namely, these are the target values S and the weighting values W. These values are entered into the computer before the optimization is carried out, for example using the keyboard 102 in FIG. 11.

It can be seen from this example that the target value for the stack of mixture 16, for the fineness of mixture 4, for the color value a 1, for the color value b 3 and for the price of the mixture is 0, since the target price is as low as possible being held. In terms of weighting, the stack, the fineness and the color value b all received the same weighting of 1. In contrast, the color value a has a Weighting 0, since in this example all color values a have the same value 1, so that a change in the color value a of the mixture cannot be achieved with this mixture, because changes in the percentages of the individual components do not lead to a change in the color value of the mixture. Therefore, the weighting is completely irrelevant here and is given as zero. The weighting of the price has also been intentionally set relatively low, to prevent the calculator from overvaluing the price. If one did not use such a trick, there would be a great risk that the computer program would lead to an excessive proportion of the inexpensive component x5, with large compromises in the other technical values, which are ultimately decisive for the marketability of the fiber product.

As explained above, the computer program endeavors to keep the loss function as small as possible and actually to keep it zero. In this endeavor, the proportions for the individual components are calculated, which are given under the name "mixture vector C1". It is striking here that the information on the proportion of component x3 is negative, which would not be possible in practice, because in order to achieve this, one would have to subtract a quantity of x3 from the mixture, which would not be sensible or hardly feasible.

The operator is therefore forced to set the value zero for x3, because no percentage of this component is absolutely necessary. With this information, the computer carries out post-optimization, ie corrects the calculated values. Here, too, the computer strives to keep the value of the loss function as small as possible, taking into account the additional boundary conditions that x3 must be zero.

With this boundary condition, the computer comes to a corrected component division, the proportions of x1, x2, x3, x4, x5, x6 being 0.4536, 0.3101, 0.0171, 0.014, 0.0791. The values for the stack, the fineness, the color values and the price that are now printed out are also interesting for the operator. He immediately sees that the calculated properties of the fiber mixture, i.e. of the card sliver or yarn are very close to the specified target values. He also notices that the price has dropped only slightly from 2,581 to 2,631 due to the elimination of component x3.

The result of the fine tuning also shows the value of the loss function as zero. In fact, this value is not zero, but simply so low that it is not displayed in the program used. In an attempt to achieve a specific value for the loss function, which is then more suitable for comparison purposes, the same example was calculated again, the weighting all being increased by a factor of 1000. The result of this variant is then shown in Table II of FIG. 16. Here you can see that with otherwise unchanged information in the upper and middle part of the table, the loss function is now shown with the value 0.135.

The other part of Table II shows the result of a post-optimization, which has been carried out differently from the information in Table 1. Here, too, it is necessary to specify the component x3 as zero. It has also been decided that the same proportions of components x5 and x6 should be present, since there is relatively little of these components left, and the spinning mill management wants to ensure that this remaining stock is fully used and will disappear from the inventory at the same time. It has also been decided not to add any components of component x4 here because this component is temporarily unavailable. All of these additional boundary conditions lead to a worse result being achieved in the post-optimization, although one can still speak of an optimization, since under the given boundary conditions the result of the post-optimization may be an optimum. The operator immediately sees that the stack of the mixture now deviates relatively far from the target value of 16 at 16.66. There is also a relatively large deviation in the fineness of the mixture. There is no deviation for the color value a, which is also to be expected since all components have the same color value a. The deviation in the color value b is not particularly pronounced. Of particular economic interest, however, is that the price for the mixture is now 2,768 instead of the previous optimum of Table I of 2,631, so that a significant deterioration can also be felt here. The value of the loss function has now risen to 0.548, which confirms that there are considerable deviations from the target values.

If the user agrees with the values shown to him, he can forward the information on the mixture proportions to the controller for the mixing ratios by entering a corresponding command, for example "confirmed", these proportions then being decisive for the corresponding target values for the individual components. The tables thus indicate the content of the user dialog. Each user dialog is very similar to the diagram in FIG. 11, with the exception that it is not necessary to specify a desired component division here, although it is entirely possible to set certain values for certain components, as in connection with the Tables has been explained. With this variant, the defined proportions of certain components represent boundary conditions for the calculation.

Although, for the purpose of illustration, the result of the initial optimization gives a negative value for the component x3, it is entirely possible to program it in such a way that the computer program always sets such negative values to zero and carries out the optimization again. This would then lead to the result of the re-optimization of Table I becoming the result of the optimization, after which the operator can, if desired, enter further boundary conditions if the values determined by the computer do not suit him for certain reasons.

Claims (21)

1. A method for mixing textile fibers, in which different types of fibers from fiber bales of different provenance are removed and mixed, characterized in that I) at least the following information is specified in a regulation: a) the initially roughly estimated desired quantitative component distribution, b) the properties of the fibers of the individual components, and c) the desired properties of the card sliver or yarn produced from the fiber mixture; II) from these specifications, the control system calculates a component distribution according to a predetermined control algorithm that comes close to the predetermined component distribution and fulfills the card sliver or yarn properties, III) the control controls the operation of a mixer mixing the individual components in such a way that the calculated component division is obtained in the fiber mixture supplied by the mixer. 2. The method according to claim 1, characterized in that in the case of feature I) additionally d) at least one control priority is specified, in the sense that compliance with at least one component component or a card sliver or yarn property has priority. 3. The method according to claim 2, characterized in that a weighting is also specified for each rule priority listed. 4. The method according to claim 3, characterized in that the weighting is predetermined by the order of the information. 5. The method according to any one of the preceding claims 1 to 4, characterized in that at least some of the desired card sliver or yarn properties are measured and communicated to the control during card sliver or yarn production, and that the control in the event of deviations from the specification the component distribution is recalculated in relation to the measured properties. 6. The method according to claim 5, characterized in that one additionally measures further card sliver or yarn properties in the laboratory and in the case of disturbing deviations also enters these into the control, whereby these deviations are also taken into account in the new calculation of the component distribution. 7. The method according to any one of claims 5 or 6, characterized in that the properties measured during the production of the card sliver or yarn are taken into account by the control only after corresponding averaging. 8. The method according to any one of the preceding claims 1 to 7, characterized in that the calculation of the component distribution is carried out according to the principle of minimum deviations or the minimum weighted deviations from the target specification. 9. The method according to claim 8, characterized in that the calculation of the component distribution is carried out on the principle of minimum square deviations or the minimum weighted square deviations from the target specification. 10. The method according to claim 8, characterized in that the calculation of the component distribution according to the following equation or the following control algorithm takes place by the quality criterion
J (u) = _o ∫ ^∞ { x ^T (t) Q x (t) + u ^T (t) R u (t)} dt
is minimized
in which
x (t) indicates the system deviations in the form of a vector, ie the deviations of the measured properties from the desired properties,
x ^T (t) is the transform of x (t),
u (t) is the control vector that indicates the desired component split,
u ^T (t) is the transform of u (t),
Q and R are matrices with which the individual components are weighted in x (t) and u (t). 11. The method according to any one of the preceding claims 1 to 10, characterized in that the control is used at the same time for setting a coarse cleaning unit, which is between a bale removal machine and the mixer, the setting of the coarse cleaning unit influencing the card sliver or yarn properties and thereby also the calculation of the component distribution. 12. The method according to any one of the preceding claims 1 to 11, characterized in that the control takes into account the setting of a coarse cleaning unit arranged between the bale removal machine and the mixer and the setting of any fine cleaning unit which may be present. 13. The method according to claim 11, characterized in that the control also serves to set at least one fine cleaning unit switched on after the coarse cleaning unit, which likewise influences the card sliver or yarn properties and thereby also the calculation of the component distribution. 14. The method according to any one of the preceding claims, characterized in that when changing the assortment, the control system coordinates the readjustment of the component distribution and the can change at the card exit, so that the transition from one assortment to the next takes place without any significant interruption and with minimal product loss. 15. A process for mixing textile fibers, in which different types of fibers from fiber bales of different provenance are removed and mixed, characterized in that I at least the following information is entered into a computer: a) the properties of the fibers of the individual components, and b) the desired properties of the card sliver or yarn produced from the fiber mixture; II a computation of components is calculated from these specifications of the computer in accordance with a predefined calculation algorithm, which at least approximately fulfills these with a minimal deviation from the desired card sliver or the yarn properties, and, if necessary, a correction of the computation of components calculated taking into account any boundary conditions or parameters also entered into the computer Makes special requests and calculates a correct component allocation, III the component division or corrected component division determined by the computer is used for setting or regulating the feeding of the individual components on the mixer in order to obtain the calculated and possibly corrected component division in the fiber mixture supplied by the mixer. 16. The method according to claim 15, characterized in that the computer itself regulates or controls the component feed. 17. The method according to claims 15 or 16, characterized in that a predetermined quantity for at least one specific component of the mixture is entered into the computer as a boundary condition. 18. The method according to any one of the preceding claims 15 to 17, characterized in that for at least some of the desired properties of the fiber mixture, a weighting is entered into the computer, which at the calculation of the component distribution or the corrected component distribution is taken into account, for example when calculating the minimized deviation in the form of a loss vector. 19. The method according to any one of claims 15 to 18, characterized in that the price of this component is entered into the computer for each component of the fiber mixture and that the control algorithm takes into account the costs of the individual components in order to minimize the total costs of the calculated component distribution. 20. The method according to claim 19, characterized in that the price of each component is an adjusted price which takes into account the real price of the component per unit of weight after removal of the dirt content from the purchased material. 21. The method according to claims 19 and 20, characterized in that the calculation of the component division is carried out by using the following equation
v (c) = (QK ^* - q) ′ ^* W ^* (QK ^* c - q) + w ^* c ′ ^* p ^* c
taking the constraint into account
g (c) = 0 = k + e ′ ^* c
as well as the constraint
0 = <ci = <1 minimized for the integers i = 1 ... n, where the symbols have the meanings given in the description.

Images