EP0477589A1 - Method for the correction of a determined measuring signal for the mass of a fiber band at an autolevellor for fiber bands with an outlet measuring device - Google Patents

Abstract

A regulating drawing frame has measuring members for the fibre quantity running through, both at the entrance and at the exit. A partially self-adjusting (adaptive) system is provided in order to allow for measurement errors and thereby achieve an improved self-adjusting optimisation of the control or regulating parameter. <IMAGE>

Description

The invention relates to a regulating drafting system, i.e. a drafting system, in which the draft can be controlled or regulated in order to even out fluctuations in mass in a stretched fiber structure. Such drafting systems are often used in so-called regulating lines in short fiber spinning, but can also be used on cards, combing machines and combing preparation machines in short fiber spinning. The same principles are of course also suitable for use in long-staple spinning.

State of the art

The principles of control technology have been used in regulating drafting systems for several decades. This made it possible to continuously improve the quality of the fiber slivers processed (provided this quality is determined solely by the uniformity of the mass per unit length).

Over the same period, intensive efforts have been made to clearly define the term "quality" in relation to the uniformity of the sliver. These efforts have led to generally accepted test procedures with the consistent provision of suitable test equipment.

With the help of the technology previously used, together with a quality-oriented organization of the spinning mill, it is now possible for every spinning mill to avoid or correct most (relatively gross) errors and to produce fiber dressings of good average quality.

Due to continuously increasing quality requirements, it is now necessary to increase this good quality level again. However, this leads to a technical area where it is no longer sufficient to apply the basic principles of control technology or the basic principles of statistical quality control in spinning. In order to achieve a further significant quality improvement, it is now necessary to take a closer look at the more intimate interactions of the measuring principles used, control or regulation principles, drive systems, distortion forces and material properties. It is always necessary to keep an eye on the principles of uniformity testing for fiber dressings, which are already defined by standards.

Quality control in the spinning mill is nowadays largely carried out in the laboratory ("off line"). For this purpose, samples are taken from the processing line, carried to the laboratory and checked. The test results are intended to enable conclusions to be drawn about the setting of the machines or the adaptation of the material to be processed to the requirements for the desired end product.

In the laboratory, (off-line method) there is time to analyze the different information, to make an appropriate interpretation of the different results and to determine the corresponding conclusions. If an attempt is made to use such methods "on line" in normal operation, where corrective intervention should be made in the process on the basis of measured values that have just been determined, it may come as no surprise that there is a great risk of a mistake or fallacy. The control system "interprets" the existing measurement data incorrectly and interferes with the process accordingly.

A first attempt to overcome this problem can be found in European patent 176 661 (US Pat. No. 4,653,153). According to this proposal, short-wave fluctuations in mass of the inlet fiber structure are compensated for by a control determining the warpage. Two control parameters, namely the gain and the time shift, can be adjusted in a controlled manner. The results of the controlled changes in warpage are determined by monitoring in the outlet of the drafting system, so that the two control parameters mentioned can be optimized on the basis of the monitoring of the results. From the point of view of control technology, there is nothing to be said against this proposal. However, it is not sufficient for the desired improvement in quality because it does not take into account measuring, processing, and technological problems. In addition, it is based on the evaluation of instantaneously obtained measured values with an intervention in the method, which takes place either immediately or after a simple time shift. The "history" of the process is ignored. Similar ideas can be found in CH-PS 672 928 (US-PS 4,819,301).

A further proposal for a "deeper" monitoring of the method can be found in EP 340 756. According to a first variant of this proposal, limit values for the signal delivered by the outlet measuring element are to be determined, an alarm being triggered or the machine being switched off when a limit value is exceeded . In this case, the product (the sliver supplied) should be checked by personnel. Depending on the results of this check, measurement errors or control errors should be concluded.

A second variant of the same proposal provides for the determination of limit values for the actuating signal which determines the delay, and likewise if one exceeds Limit value an alarm is triggered or the machine is switched off. In this case, the sliver should be checked by the staff, depending on the test results for errors in the inlet measuring system or in the production of the original material (ie in the processing machines in front of this drafting system).

Monitoring the measurement signal of the outlet measurement system can convey certain information about malfunctions. However, this measure alone is certainly not enough to achieve a significant improvement in quality. The monitoring of the control signal proposed in EP 340 756 has hardly any advantages in combination with an alarm or when the machine is switched off. By the time the staff checked it, the defective fiber sliver had long been processed (corrected) by the drafting system, so that important information regarding the error is no longer available. Because the monitoring is set to react only to a short-term (possibly rare) "outlier", the piece of fiber sliver to be examined by the staff no longer contains a corresponding "event", so that the risk of a fallacy again arises.

Our own Swiss patent application No. 2955/89 dated August 11, 1989 describes a further developed system, according to which in particular measurement problems in determining the high-frequency portion of the mass fluctuations (in the inlet) can be better taken into account.

It is the object of this invention to further develop the regulating drafting device in such a way that the interactions which are decisive for its function are taken into account even better than in the above-mentioned Swiss patent application.

The invention

The invention provides a regulating drafting device for fiber slivers, which is equipped with an outlet measuring element, at least one draft zone, a drive system and a control or regulation for the drive system. The open-loop or open-loop control responds to a measurement signal supplied by the outlet measuring element so that the warping in the warping field mentioned is changed via the drive system in such a way that mass fluctuations in the sliver are corrected.

The invention is characterized in that the measurement signal of the outlet element can be adapted as a function of the operating conditions in order to compensate for effects on the measurement result caused by these conditions.

One such operating condition is the amount of warpage that was exerted on the sliver that caused the measurement signal. Another such operating condition is the delivery speed.

Fig. 1

zeigt ein Streckwerk mit einem Vor- und Hauptverzugsabschnitt und den prinzipiellen Messeinrichtungen nach unserer schweizerischen Patentanmeldung Nr. 2834/89 vom 31. Juli 1989,

Fig. 2

zeigt einen Messwandler für das Einlaufmessorgan 9.1,

Fig. 3

zeigt das Funktionsprinzip eines Verfahrens nach unserer schweizerischen Patentanmeldung Nr. 2955/89,

Fig. 4

zeigt eine vereinfachte Version eines Verfahrens nach Fig. 3, wobei der Rechner in Vordergrund gestellt wird,

Fig. 4A,B

sind Diagramme zur Erklärung der Auswertung des vom Einlaufmessorgan gelieferten Signals, und

Fig. 5

zeigt eine Anpassung des Systems zur Berücksichtigung allfälliger Messfehler im Auslaufmessorgan.

The method and exemplary embodiments of the device according to the invention are explained in more detail with reference to the following figures, systems for the time being being described as the starting point, in accordance with our earlier patent applications:

Fig. 1

shows a drafting system with a preliminary and main drafting section and the basic measuring devices according to our Swiss patent application No. 2834/89 dated July 31, 1989,

Fig. 2

shows a transducer for the inlet measuring element 9.1,

Fig. 3

shows the principle of operation of a method according to our Swiss patent application No. 2955/89,

Fig. 4

3 shows a simplified version of a method according to FIG. 3, the computer being placed in the foreground,

4A, B

are diagrams for explaining the evaluation of the signal supplied by the inlet measuring element, and

Fig. 5

shows an adaptation of the system to take into account any measurement errors in the outlet measuring element.

Fig. 1 shows a schematic representation of an embodiment of the route. Several slivers 15.1 - 15.6, six in the example, are guided side by side through several roller systems 1 - 6. Due to the fact that the circumferential speed of the rollers increases in two stages in the direction of transport of the fiber material, this is pre-drawn over the first stage (pre-drafting) and further drawn over the second to the desired cross-section (main drafting). The fleece 18 emerging from the route is thinner than the fleece of the fed strips 15.1 - 15.6 and correspondingly longer. Because the warping processes can be regulated depending on the cross-section of the fed tapes, the tapes or the fleece are made more uniform as they pass through the route, i.e. the cross-section of the emerging fleece is more uniform than the cross-section of the fed fleece or tapes. The present route has a pre-drafting area 11 and a main drafting area 12. Of course, the invention can also be used in an analogous manner in connection with routes with only one or more than two delay areas.

The belts 15.1 - 15.6 are fed into the line by two systems 1 and 2 of conveyor rollers. A first system 1 consists, for example, of two rollers 1.1 and 1.2, between which the fed belts 15.1 - 15.6, which are combined to form a loose fleece, are transported. In the transport direction of the belts follows a roller system 2, which here consists of an active conveyor roller 2.1 and two passive conveyor rollers 2.2, 2.3. During the feed through the roller systems 1 and 2, the fed strips 15.1 - 15.6 are brought together to form a composite 16. The peripheral speeds v 1 and v 2 (= v _in ) of all the rolls of the two roll systems 1 and 2 of the feed are approximately the same size, so that the thickness of the fleece 16 essentially corresponds to the thickness of the fed strips 15.1 - 15.6.

The two roller systems 1 and 2 of the feed are followed in the transport direction of the fleece 16 by a third system 3 of pre-drafting rollers 2.1 and 3.2, between which the fleece is transported further. The peripheral speed v₃ of the pre-drafting rollers is higher than that of the infeed rollers v _1.2 , so that the fleece 16 is stretched in the pre-drafting area 11 between the infeed rollers 2 and the pre-drafting rollers 3, its cross section being reduced. At the same time, a pre-warped fleece 17 is created from the loose fleece 16 of the fed-in belts 17. The pre-warping rollers 3 are followed by a further system 4 of, for example, an active conveyor roller 4.1 and two passive conveyor rollers 4.2, 4.3 for further transport of the fleece. The peripheral speed v₄ of the conveyor rollers 4 for further transport is the same as v₃ of the pre-drawing rollers 3rd

The roller system for further transport 4 is followed by a fifth system 5 of main drafting rollers 5.1 and 5.2 in the transport direction of the fleece 17. The main drafting rollers in turn have a higher surface speed v₅ than the preceding transport rollers 4, so that Pre-drawn nonwoven 17 between the transport rollers 4 and the main drafting rollers 5 in the main drafting area 12 is further drawn to the finished warped nonwoven 18, the nonwoven 18 being brought together into a band via a funnel T.

Between a pair 6 of outlet rollers 6.1, 6.2, the circumferential speed v = (= v _out ) of which is approximately the same as that of the preceding main drafting rollers (v₅), the finished stretched belt 18 is guided out of the path and placed, for example, in rotating cans 13.

The roller systems 1, 2 and 4 are driven by a first motor 7.1 via a gear or preferably via toothed belts. The pre-drafting rollers 3 are mechanically coupled to the roller system 4, the translation relative to the roller systems 1 and 2 being adjustable or a setpoint being predeterminable. The gear (not visible on the figure) determines the ratio of the peripheral speeds of the inlet rollers (v _in ) and the peripheral speed v₃ of the pre-drafting rollers 3.1, 3.2, hence the pre-drafting ratio. The inlet rollers 1.1, 1.2 can also be driven by the first motor 7.1 or by an independent motor 7.3.

The roller systems 5 and 6 are in turn driven by a second motor 7.2. According to the invention, the two motors 7.1 and 7.2 each have their own controllers 8.1 and 8.2. The regulation takes place via a closed control loop 8.a, 8.b or 8.c, 8.d. In addition, the actual value of one motor can be transmitted to the other motor in one or in both directions via a control connection 8.e so that everyone can react accordingly to setpoint deviations of the other.

The total cross-section of the fed strips 15.1 - 15.6 is measured at the inlet of the line by an inlet measuring element 9.1. At the exit of the line, the cross section of the emerging strip 18 is then measured by an outlet measuring element 9.2.

A central computer unit 10 transmits an initial setting of the target size for the first drive 7.1 via 10.a to the first controller 8.1. The measured variables of the two measuring elements 9.1, 9.2 are continuously transmitted to the central computer unit via the connections 9.a and 9.b during the stretching process. From these measurement results and from the setpoint for the cross section of the emerging belt 18, the setpoint for the second drive 7.2 is determined in the central computer unit and any other elements using the method according to the invention. This setpoint is continuously given to the second controller 8.2 via 10.b. With the help of this control system, fluctuations in the cross-section of the fed strips 15.1 - 15.6 can be compensated for by appropriate regulation of the main drafting process or the strip can be made more uniform.

Position regulators (not speed regulators) are used as regulators in the context of the auxiliary regulation, since they also guarantee regulation when the motor is at a standstill. The corresponding controllers 8.1, 8.2 (or any other controllers within the scope of the design variants) can contain separate computer units (for example with digital computing elements; microprocessors) or can also be designed as a module of the central computer unit 10.

The measuring principle will be explained in more detail below. In the illustrated embodiment of a controlled route, a constant early delay should take place. The regulation of the strip cross-section or its Uniformity is essentially achieved by changing the delay in the main delay area 12. The inlet measuring element 9.1 supplies the measuring signal on the input side with the information about the cross section of the fed strips 15.1-15.6.

Obtaining the desired run-in measurement signal is known to present measurement difficulties. A cross-sectional measurement without impairing the material and with high dynamics is difficult to do in the conventional way. As a consequence, an indirect measurement procedure must be carried out with a transducer. Various conventional converters only provide insufficient results for the desired purpose. A measuring capacitor 21 according to FIG. 2 is therefore used in connection with this invention, through which the fed strips 15.1-15.6 run. This takes advantage of the fact that the fiber mass of the strips between the capacitor plates, which fluctuates as they pass through, acts as a change in the dielectric.

When these bands pass through the capacitor 21, with an alternating current U applied, a conclusion can be drawn about the dielectric, for example by measuring the voltage U across the capacitor. However, it must be taken into account that the moisture content of the tapes and other disturbances can severely affect the measurement. With respect to the moisture content, the dielectric constant ε w of water is 81 compared to the dielectric constant of, for example, cotton ε b, which is approximately 4. In other words, the difficulty lies in obtaining the desired signal directly from the transducer via the fiber mass present in the capacitor at a certain point in time.

The voltage U is measured across the capacitor and the signal obtained in a real part R _x and Imaginary part C _x split. As will be explained further below, these signals R _x and C _x are evaluated as part of the regulation, the outflow measurement signal being used in the process. One of the reasons for the difficulties with the input-side measurement is that the control is designed in such a way that measurement errors are compensated for as part of an adaptive control.

The outlet measuring element 9.2 can be a conventional measuring instrument which delivers a signal A _out with the information about the cross section of the emerging belt 18. This signal is also subsequently used for the control. It should be noted that the required measurements can not only be carried out directly at the inlet and outlet, but it is only necessary that one measuring element is arranged before and one after the controlled system (in the control-technical sense), ie here the main warping area 12. With regard to a favorable time dependency of the regulation, it would also be advantageous, for example, to arrange the input-side measuring element directly in front of the main delay area 12.

It is assumed that both high and low frequency changes or unevenness of the belt should be corrected for an optimized control. The regulation is intended to keep the mean value of the strip as constant as possible (first priority) and to regulate unevenness. The relevant deviations in the controlled variable can be recorded as high- and low-frequency components of the measured controlled variables as part of the control. In terms of measurement and control technology, there is the problem of obtaining information about these variables and converting them into the desired manipulated variable. Especially in the case of high-frequency changes, the runtime between the measuring and control element must be taken into account. On the input side, ie with the inlet measuring process 9.1, there is the option of gain high-frequency signal components. Because of the dead time of the measurement on the output side by means of the regulation dependent on the outlet measuring element 9.2, only the low-frequency components of the signal can be compensated in the context of the regulation. Problems and errors caused by measurement technology are now also taken into account according to the invention in the context of the control by taking the measurement signals of the outlet measuring element 9.2 into account in order to adapt the control to measurement errors on the inlet side or other deviations. According to the invention, a map R, which is preferably determined empirically and continuously adapted during operation, is provided.

FIG. 3 illustrates the control principle and the method according to the invention in a schematic overview of the main control. The distance is indicated by arrows which indicate the direction of travel of the belt, as well as by two blocks for the preliminary draft 11 and the main draft 12. The actual cross section m _{E of} the belts at the inlet is represented by the size m _e , the actual cross section m _{A of} the finished warped belt by the size m _a . The belts are fed _in at the speed v _in at the inlet and the finished belt exits at the speed v _out . The size of the pre-draft K1 can be adjusted by means of a specification element 19. The controlled system (in the technical control sense) is formed here by the main delay area 12. The running time between the inlet measuring element 9.1 and the main drafting area 12 is identified by t1, that between the main drafting area and the outlet measuring element 9.2 by T2. The measured variables A _out , R _x and C _{x of} the measuring elements 9.1, 9.2 represent input variables of a control system. This contains a central computer unit 10 which contains the measured variables C _x , R _x , the temperature I _T and any further information I _1-n , such as humidity, air pressure, etc. are fed. The size A target _is specified as the reference variable.

For the sake of clarity, the control system is divided into several "paths" 1 - 4. A first path 1 contains the central computer unit 10 with inlets and outlets and a plurality of time elements Z1.1-Z3 and, according to the invention, is used to prepare the measurement data. A second path 2 is used to optimize the delay time t1. A third path 3 serves to keep the band mean value constant and to compensate for long-term disturbances. Finally, a fourth path 4 is provided, which provides for an optimized compensation of short-term disturbances. It is anticipated that digital control is preferably used in the context of the invention. This makes it possible to implement all elements of the control system in one computer. To illustrate the control principle, the essential elements necessary for the explanation of the invention are schematically broken down in FIG. 3.

Beginning at path 3 (constant of the average value) is a comparator 35 is provided the formation of a difference between the outlet signal A _out and the target value A _is performs. The deviation dA determined in this way is fed via an I-link 38 to an addition point 36. The signal △ m is formed by integrating the mean value deviations in an I element 38. In the addition point 36, the signal △ m is supplemented by the addition of 1. In a second addition point 37 these deviations and the deviations △ h caused by short-term disturbances, which are determined in path 1 and 4 according to the following explanations, are added and finally the factor 1 + △ m + △ h in a multiplication point 39 with the predetermined nominal value K3 of the main default multiplied. The corresponding Multiplication gives the required manipulated variable y for the regulation of the main delay.

The outflow measurement signal A _out is further fed to a high-pass element 47 of path 2. The filtered signal is squared at a multiplication point 40 and the signal △ H is obtained therefrom, which indicates the high-frequency portion of the mean value fluctuations. The high-frequency components are taken into account for this path, which in this exemplary embodiment are up to approximately 300 Hz. The signal △ H is fed to a first control element R1 with a transfer function to minimize △ H. The output of the control element R1 forms the signal S _t1 , which optimizes the delay time of various time elements Z1.1, Z1.2, Z4 and is fed directly to the central computer unit (10).

A map element 50 is provided as the connecting core of paths 1 and 4. This can be designed, for example, as a readable and readable memory and can itself be integrated into the central computer unit 10. An empirically determined output map R with respect to the sizes R _x and C _{x is} stored in this map element and relates to the size m _e = f (R _x , C _x ). The measured value pairs R _x , C _{x are} supplied to the map element 50 and this supplies the quantity m _e as the output signal. The map fR is continuously adjusted during operation. This adaptation takes place in path 1. In this exemplary embodiment, the signals R _x , C _x are fed into the central computer unit 10 with a delay in corresponding timing elements Z1.1-Z2.2. The timing elements Z1.1 - Z2.2 serve to take into account the total running time t1 + T2 from the inlet to the outlet measuring element. The filtered variable m _{e (t1)} , delayed taking into account the running time t1 and adjusted for delay in a division element 43, is sent to a further input of the central computer unit via a timing element Z3 fed. The signal A _out with the information about the outlet band cross-section m _A , represented by the measured quantity m _a , is preferably also filtered before it is fed to the central computer unit 10, the low-frequency signal components being trimmed in a corresponding filter 46 of the path 1. Instead of using these timers, the transit time t1 can also be taken into account directly by the central computer unit by supplying the output signal S _t1 of path 2 to it.

All of the signals supplied to the computer unit are used in the following for cleaning the map R of the map element 50, in that the output of the computer unit 10 into the map element 54 contains the (effective) size m _e for the respective value pair C _x , R determined by evaluating the measurement data _{x is} transmitted. This ensures a permanent adaptation of the map R to changes within the control process. It can be seen that the central computer unit 10 must at least evaluate the signals m _e , R _x , C _x and m _{a in} order to ensure the map adaptation. The mentioned additional measurement data I _T , I _1-n can, however, bring about a further improvement in the control under certain conditions.

Similar to path 2, path 4 provides for filtering the signal A _out , but this time with a bandpass element 48 instead of a highpass element. The bandpass element 48 is followed by a multiplication point 44 and a control element R2 for minimizing the corresponding signal B. The control element R2 provides at its output a factor f _b , which is linked in a multiplication point 42 with the signal m _{e (t1)} . This signal m _{e (t1)} is present at the output of a filter 49, to which the signal m _e from the map element 50 is supplied via a timing element Z4. This filter 49 cuts the low-frequency signal components. Path 4 further contains a threshold switch 25 with an adjustable default value δ on both sides of an average value. If the signal m _{e (t1)} lies below this preset value δ, ie within the tolerance limits around the mean, the switch is in a first position p1. As soon as the specified value δ is exceeded in one direction or the other, ie large fluctuations of m _e around the mean value, the switch switches to a position p2 in which the signal m _{e (t1)} is looped directly to path 3, so that these fluctuations are fully taken into account for the main default.

However, if the values for m _{e / t1} are below this default value δ, path 4 is optimized. The signal _{me (t1)} is multiplied in the multiplication point 42 by the factor f _B determined by means of the minimization function of the control element R2 and the output signal of the multiplication point is fed to the path 3 via the switch 25. The switchover by means of the threshold switch 25 and the consideration of the optimization by the control element R2 prevents that, in the case of small and very small, short-term mean value deviations, any interference, for example caused by noise, is introduced into the path 3.

At the same time, the threshold switch is used to switch the optimization on or off using the control elements R1, R2. If m _e lies above the preset value δ, the optimization of the control elements R1, R2 is switched off, otherwise it is switched on. It is not absolutely necessary to switch off the respective optimization brought about by the control elements R1, R2 when the preset value δ is exceeded, since the corresponding regulation can also run away by means of compensation elements. In the context of digital regulation, however, it is very easy to switch the corresponding regulations on and off possible, so this variant is preferred. After the optimization R1 is switched off, the time delay t1 last set remains unchanged until the optimization R1 is switched on again.

The threshold switch can also be implemented by a non-linear element or can be integrated in the characteristic map R. In the latter case, in addition to the output variable m _e , the map element 50 also supplies the signal required for activating or deactivating the optimization of the control elements R1, R2 or an amplitude-dependent parameter.

In the present exemplary embodiment, the high-pass element of the path 2 can, for example, filter frequencies above 100 Hz, the band-pass filter those in the range from 10-100 Hz. The frequency ranges depend on the throughput speed of the bands, which in the above information are in the range around 600 m / min. was accepted. The frequency ranges can also be adjusted based on the delivery speeds.

It must be noted that transfer functions of the control elements R1, R2 can vary depending on the configuration of the control system. In a preferred embodiment of the invention, the filters of paths 2 and 4 can be omitted and instead the transfer functions can be determined in such a way that the relevant frequencies are taken into account as required. Of course, the filter 46 of the path 1 can also be omitted and the filtering can be implemented in the context of the central computer unit 10. Due to the possibility of changing the parameters of the corresponding transfer functions, there is also the advantage that an adaptation to different operating conditions (eg variable throughput speed of the tapes) can be carried out easily.

In this sense, a special embodiment provides for an adaptive adaptation of the control parameters. The parameters of the transfer functions of the control elements R1, R2 are changed in the course of the control, so that the variation of the manipulated variable is minimized. In such an embodiment, the parameters of the transfer functions are determined by the central computer unit 10 from the measured variables. With adaptive control, great value must be placed on stability, which is achieved by defining key data of the map.

The central computer unit 10 is preferably implemented by a digital computing element. It is obvious that the functions of the different paths 1 - 4 in FIG. 3, which are explicitly illustrated to explain the principle of the method, can be partially or completely integrated in a uniform computer.

The output map R for m _e can be determined, for example, by static measurements on the measuring capacitor 21 and then stored in tabular form. It should be noted that different characteristic diagrams must be determined for modified measuring methods. The principle according to the invention can therefore also be carried out with corresponding characteristic diagrams for other inlet and outlet measuring elements.

The control principle according to the invention ensures a very good uniformity even in the event of unforeseen changes in the operating conditions. In particular, measurement errors on the inlet side are also compensated as part of the control. Both short-term disturbances and slow changes can be optimally compensated within the scope of this regulation. If the described method for the main control of the drafting system is combined in connection with the auxiliary control of the independent drive groups and a correspondingly meshed control is provided, particularly favorable conditions result. The through the Main control determined control variable y is accordingly used as a setpoint for the controller 8.2 of the drive for the main warping area 12.

For the sake of completeness, it should be mentioned here that the method according to the invention is suitable for all devices of the textile industry which require a regulation of a stretching process and are not limited to the distance mentioned in the description.

Various operations of the arrangement according to FIG. 3 will now be explained in more detail with reference to FIG. 4 together with the associated diagrams FIGS. 4A to 4C. 4 shows a simplified version of FIG. 3. FIG. 4 again shows the inlet and outlet measuring elements 9.1 and 9.2, the pre-delay field 11, the main delay field 12 and the computer 10 with its inputs for the signals R _x , C _x , A _should , A _out , 1 _T and 1 _in . This figure highlights the fact that all rule operations are implemented in the software of the computer, ie the "elements" of paths 1 to 4 of FIG. 3 are aspects of the programming of the computer 10.

• 1. die Bildung bzw. Anpassung des Kennfeldes, und
• 2. die entsprechende Aufteilung des Ausgangssignals vom Messorgan 9.1 in seinen Komponenten R_x und C_x.

The operations that will now be explained in more detail are:

• 1. the formation or adaptation of the map, and
• 2. the corresponding division of the output signal from the measuring element 9.1 into its components R _x and C _x .

The size m _e should correspond to the mass of the fibers located in the measuring field of the inlet measuring element 9.1. The individual signal components R _x and C _x do not correspond to this mass because they are also dependent on at least one further variable (the "parameter"). However, it is possible to create a "family" for this important parameter of "curves" which determine the course of the two signal components R and C as a function of the fiber mass (ie as a function of m _e ) for any selected values of the parameter. This has been done in the diagrams 4A (signal component C _x ) and 4B (signal component R _x ), which is not about the exact course of the characteristics, but only about the representation of the principle. The parameter is the water content of the fiber structure, which can be represented, for example, as a percentage of the fiber mass. As examples, three characteristics are shown in FIGS. 4A and 4B - one characteristic at 10% water, one characteristic at 20% water and one characteristic at 30% water.

It is important that the signal component R changes mainly with the water content and remains practically constant when the fiber mass is changed (FIG. 4B shows horizontal "curves" in FIG. 4B, which is not quite true in practice, but can be assumed as an approximation.

This means that a signal component pair R _x , C _x can be uniquely assigned to any value of m _e (fiber mass in the measuring field). The “real” signal component R enables “selection” from the “curves” in FIG. 4A, so that the “imaginary” signal component can be used to determine the fiber mass (size m _e ) on the basis of the relevant characteristic.

Accordingly, when the characteristic diagram is formed, individual assignments of empirically determined signal component pairs R _x , C _x to known fiber masses or fiber and water masses are entered in the characteristic diagram (in a memory of the computer 10). A theoretically calculated "model" of the map then enables the empirically determined values to be extrapolated in order to form a sufficiently extensive and detailed map (as a first approximation) around the desired working area (Fiber mass or water content) with the desired precision (ie level of detail of the map). The system can now work (but not optimized) with this map, which is partly theoretical and partly empirical.

Before cleaning the map, which will be described below, the system will not work optimally, since the map, which is largely theoretically formed, will hardly meet the practical conditions. When cleaning the map, it is important to assign the signal component pairs R _x , C _x in practice to the fiber masses that apply to them and to correct the map accordingly. This is made possible by the fact that the outlet measuring element 9.2 has a different structure than the inlet measuring element 9.1 and responds directly to the fiber mass (or to the cross section of the belt). The preferred outlet measuring element is a pair of sensing rollers, for example according to our US Pat. No. 4,539,729.

When performing the cleanup (i.e. when reassigning the signal component pairs R _x , C _x to their sizes m _e ), it is necessary to consider both the time difference between the measurement in the inlet measuring element and the outlet measuring element as well as the processing of the tape between these elements (in Drafting system). It can be assumed that there is a constant advance K1 (FIG. 3) which has been entered in the computer 10 as the machine setting. At a constant delivery speed (speed of the delivery cylinder), a constant throughput time T₂ from the drafting device to the outlet measuring element 9.2 can also be assumed. In order to achieve the desired compensation of the fiber masses per unit length of the belt, it is necessary to change the speed of the inlet rollers up to the inlet into the main drafting zone, with consequent changes in the Main delay and the lead time from the inlet measuring element 9.1 to the main delay field.

3 is accordingly assigned a value for the fiber mass in the delivered band that is currently determined by the outlet element 9.2 to a pair of signal components R _x , C _x that was generated a certain time earlier by the inlet measuring element 9.1. This assignment is made possible by taking into account the time shifts t 1 and T 2. When cleaning the map, however, it is not the fiber mass in the delivered tape that is of interest, but the corresponding fiber mass in the feed tape, which can be reconstructed by taking into account the distortion that was exerted on the feed tape. This reconstructed value of the fiber mass in the feed tape is assigned (as the size m _e ) in the map to the appropriate signal component pair, the originally (theoretically) calculated value of the size m _e for this signal component pair being canceled.

After a certain period of "experience", the system has built up its own (tailored) map in this way and can work "optimized" accordingly. However, only one "disruptive factor" (the water content of the supply belt) has been taken into account. In practice, there are other predictable and unpredictable disturbances. The disturbances include, for example, changes in air humidity which influence the "stretchability" of the fiber structure and the delivery speed, which can influence the behavior of the control loop. Where major changes are to be expected, it can be advantageous to define the changeable variable as a "parameter" and to create a separate map for different values of this parameter. It then changes from one map to another depending on, for example, a signal on input 1 changed _in or from the setting of delivery speed.

However, not all influences of the various disturbances can be predicted. The aging of the inlet measuring element (or the outlet measuring element) or the stretching rollers will influence the system, so that the behavior of the system changes over time. A change in the processed material type can also lead to changes in the system behavior. Such changes can, however, be taken into account by continuously adapting the map.

It is important that the inlet measuring element delivers two signal components that can be clearly assigned to a determinable fiber mass as a pair and that the determination can be carried out on the basis of the signal supplied by the outlet measuring element and known or ascertainable parameters of the controlled system.

The distribution of the signal of the inlet measuring device 9.1 is an important aspect of the method in its real and imaginary parts. Fig. 4 therefore shows one way to effect this division.

4 comprises a capacitor 9.1, an amplifier 100 and two rectifiers 102, 104 each with a smoothing element 106 and an amplifier 108. An alternating current source Q supplies energy with a predetermined frequency to the capacitor and also to the rectifier 102, 104 wherein a rectifier 102 is in phase and a rectifier 104 is fed with a 90 ° phase shift with respect to the capacitor.

Rectifier 104 provides a signal representing 1 / R _x and rectifier 102 provides its signal to an addition element 110, where it is compared with a reference value Co is combined, which corresponds to the capacitance of the "empty" capacitor (ie without sliver). The output line of this element 110 carries the signal C _x in the form of deviations from the reference value.

The description so far has focused on the advantages of the new system that arise in connection with the "admixing" of a disturbing material (water) in the fiber structure. But even if the fiber structure to be scanned only contains fibers (and air), the new system can offer an additional advantage, namely if the inlet measuring element cannot provide absolute values in itself but only relative results. This is e.g. the case where the inlet measuring element is formed as a measuring capacitor (Fig. 2).

Such a measuring device is very practical where it is necessary to monitor fiber material in the form of a fleece. This type of measuring device has so far had the disadvantage that it could not specify the amount of fiber itself but could only indicate fluctuations in the amount of fiber. By calibrating the inlet measuring element on the basis of the measured values of the outlet measuring element, it is now possible to obtain absolute values for the fiber mass (in the form of the size m _e ) from the inlet measuring element. The prerequisite for this is, of course, that the map has been adapted to the given conditions.

The outlet measuring element 9.2 plays an important role both in the control (path 3, Fig. 3) for keeping the mean value of the delivered tape constant and in the control (paths 1, 2 and 4, Fig. 3) for compensating for short-wave fluctuations in mass. While the inlet measuring element has to react to a “fleece” (FIG. 2), the outlet measuring element can be placed at a measuring point after the stretched nonwoven has been combined into a band (in funnel T, FIG. 1).

It is therefore possible to provide a pair of sensing rollers in the outlet, which reacts to the cross-section of the tape supplied. A pair of sensing rollers is practically insensitive to the water content of the belt. This type of transducer is therefore ideally suited to suppressing measurement errors caused by moisture in the inlet measuring element via the map. A measuring sensor that reacts to the strip cross section is subject to its own measuring errors, which must also be taken into account. The main error in measuring the cross-section of the strip is caused by the air entrained by fibers inside the strip.

It is important to recognize that a constant measurement error does not cause problems for either the control or the regulation. Difficulties only arise if the measurement error is subject to unpredictable fluctuations, which then lead to incorrect reactions of the control system.

Fluctuations in the amount of entrained air are due to changes in the packing density of the fibers in the belt, ie changes in the space between the fibers. The amount of air in the fiber structure during measurement also depends on how much air was squeezed out when the fiber structure was compressed. Both the packing density and the resistance which resists the squeezing of air depend essentially on the degree of parallelization of the fibers. The degree of parallelization of the fibers in a belt supplied by a drafting system depends on the one hand on the degree of parallelization of the fibers in the feed material and on the other hand on the distortion exerted in the drafting system. Since the warping in a regulating drafting unit changes constantly, the packing density of the fibers in the delivered belt can be changed and therefore a variable amount of air in the belt cross-section can be expected.

An even stronger effect can be found in the event of a belt break in the inlet. This leads to a corresponding decrease in the warpage and therefore to a lower degree of fiber parallelization. There is more space in the fiber structure to carry air with it, and the pair of sensing rollers squeezes a lower proportion of the air carried along out of the fiber structure. The measurement of the "amount of fibers" in the pair of feeler rollers (which reacts to the cross section of the "fiber structure") is accordingly too high.

Accordingly, a correction of the output signal of the outlet measuring element should be carried out at least in the event of high distortion changes in order to eliminate the effects of the variable amount of distortion on the signal level. This can easily be carried out by the computer 10 (FIGS. 3 and 4) because this computer receives the corresponding information about the amount of delay (with a suitable time shift, link Z3, FIG. 3).

It will hardly be necessary to adapt the signal A _out to every small change in delay. However, at least one adjustment can take place if the original material is subjected to a relatively high delay, for example after a belt break in the inlet. FIG. 5 schematically shows a "hardware solution" which could also be implemented by programming the computer 10. This solution comprises a "signal transmitter" 120 (eg element 43, FIG. 3), which delivers a signal dependent on the delay to a time shift element 122 (eg element Z3, FIG. 3). The time-shifted signal is forwarded to a threshold value element 124, so that when the threshold value, which corresponds to a predetermined amount of delay, is exceeded, a correction signal is passed on to the addition point 128 via an amplifier 126.

The addition point 128 also receives the output signal A _{out of} the outlet measuring device 9.2 and sums this signal with the correction signal when the latter is received by amplifier 126. The result is passed on to paths 1, 2, 3 and 4 via output 130 for evaluation. If no correction signal is generated because the threshold value determined by the element 124 has not been exceeded, the addition point 128 passes the signal A _out on via the output 130 without correction.

As already mentioned, the degree of parallelization of the fibers in the delivered tape depends not only on the amount of warping applied but also on the degree of parallelization of the fibers in the original material. This increases steadily in the spinning line from the card to the final spinning process. The earlier the regulating drafting device is used in the processing line, the more important it is to correct the output signal of the outlet measuring element - and the appropriate correction also depends on this "environmental factor". It is therefore advantageous in practice to make the amplifier 126 adjustable so that the correction signal can be adapted at this point in the processing line.

It is not absolutely necessary to provide the threshold switch 124. A signal dependent on the delay height can be continuously supplied to the amplifier 126, so that the signal A _{out is} always corrected as a function of the delay height. Two or more threshold values could also be defined that produce different corrections, the corrections being known empirical values.

Another problem in the evaluation of the output signal from the outlet measuring element arises from the aforementioned squeezing of air during the measuring process itself. This effect also depends on the delivery speed. The higher the delivery speed, the less air can be pressed out by the pair of rollers. An increase in Delivery speed therefore results in an apparent decrease in the strip cross-section (the fiber mass).

With a constant delivery speed in operation, there is therefore no error, but when the drafting system is stopped or running up, the measured values of the outlet measuring element are falsified. In the case of a drafting unit with a variable delivery speed to adapt the warping to fluctuations in fiber mass, incorrect evaluation can also be caused in normal operation when there are high warpage changes.

This can be avoided by correcting the output signal of the outlet measuring element as a function of the current delivery speed (at least during braking or ramp-up). The correction has not been shown individually in FIG. 5, since it can be carried out essentially as for the distortion correction shown. The correction signal is of course not derived from the delay but from the delivery speed and sent to a suitable point (e.g. the addition point 128) for combination with the output signal from the measuring element 9.2.

In connection with the outlet measuring device, automatic optimization by checking the results is not possible, because the outlet measuring device itself represents the last control over the results obtained. It is therefore all the more important to compensate for falsifications of the signal at this point. The optimization of the correction can be carried out empirically by the operating personnel.

Claims (8)

A regulating drafting device for slivers with an outlet measuring element, at least one distortion field, a drive system and a control or regulation for the drive system, the control or regulation reacting to a measurement signal supplied by the outlet measuring element in order to cause the delay in the said distortion field via the drive system change that mass fluctuations in the sliver are corrected, characterized in that the measurement signal of the outlet member is adjusted depending on the operating conditions to compensate for the effects of these conditions on the measurement results. A drafting system according to claim 1, characterized in that the measurement signal of the outlet measuring element is adapted as a function of the amount of warpage which was exerted on the fiber assembly piece which caused this measurement signal. A drafting system according to claim 1 or 2, characterized in that the measurement signal of the outlet measuring element is corrected as a function of the delivery speed. A drafting system according to claims 1, 2 or 3, characterized in that the outlet measuring element is suitable for determining the cross section of the delivered belt. A regulating drafting system with a measuring element in the inlet, which reacts both to the fiber mass passing through and to the amount of a substance carried with the fiber mass, a measuring element in the Spout that does not react to the carried material, a drive system, at least one warping field and a control system that responds to the measuring elements in order to reduce fluctuations in the fiber mass passing through the inlet measuring element up to the passage through the outlet measuring element due to distortion changes by means of obtaining a measurement signal by an inlet measuring element with two such components that component value pairs can be clearly assigned to a determinable fiber quantity, and by means of assigning component value pairs to fiber quantities determined by the outlet measuring element, taking into account the effect of the drafting device on the fiber mass between the inlet and Outlet measuring elements. A drafting system according to claim 5, characterized in that the amount of fibers determined by the outlet measuring element is only determined after adaptation of the output signal of this element according to claims 1 to 4. A method for determining the amount of fibers in the outlet of a drafting system which corresponds to an output signal from a measuring element, this signal being dependent both on the quantity of fibers and on the amount of air in the measuring element, characterized in that the output signal prior to the determination is based on the operating conditions of the drafting system is adjusted. A method for evaluating the output signals from a measuring element in the inlet of a drafting unit, this element reacting to the amount of fibers present in the measuring element, characterized in that a measuring signal from this measuring element with two such Components are obtained that component value pairs can be unambiguously assigned to a determinable fiber quantity, and that these component value pairs are assigned fiber quantities that are determined by a measuring element in the outlet of the drafting system, taking into account the effect of the drafting system on the fiber mass between the inlet and outlet measuring elements.

Images