JPS60246499A - Central detection of measured value at several measuring points - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION Industrial Field of Application The present invention relates to a method for measuring a plurality of measurement points in which the data occurring at each measurement point is transmitted repeatedly at time intervals to a central data detection and/or data processing device. Concerning a method for centrally detecting the measured values of.

BACKGROUND OF THE INVENTION In the method described in German Patent No. 6,005,746, measurement data occurring at several measuring points on a large textile machine are detected and processed by a central data processing unit. The readout speed is increased by providing a plurality of distributed data processing devices between a central data processing device and a large number of measurement locations. Only a limited number of measuring points are assigned to these distributed data processing devices.

The distributed data processing device performs data reading (scanner) and temporary storage.

Problem to be Solved by the Invention This is, on the one hand, an expensive arrangement, but on the other hand, the readout speed and readout frequency are increased. However, the above configuration has the drawback that only instantaneous values of measurements at successive 17 points in time are detected. As a result, only random values are detected and evaluated, and these random values cannot represent reliable information about the course of the Zorothes and the quality of the product obtained.

The object of the invention is to provide time-gap quality monitoring of a process at several measuring points with measuring values of the same type that are read out and processed at time intervals by a data detection device. It is to guarantee that

Measures for Solving the Problem For this purpose, only the maximum values that occur consecutively within each readout time interval are detected and retained as output signals of the measurement location until readout by a central data detection device. The retained signals are preferably erased after being read and detected by a central data processing unit. Due to the time-dependent restriction on the readout of the measured values, time-independent information on the quality of the process monitored by measurement is obtained.

Evaluation of measurement results can be further simplified and speeded up in the following manner. In other words, a predetermined bandwidth of the permissible measured value is determined not only for the permissible measured value but also at each measuring point, and when this width is exceeded, a fault signal is generated.

In a preferred embodiment of the invention, the average value of the measurement results is also determined and prepared for periodic readout. The average value of the measurement result by itself has a very high information value for quality, as does the local maximum value.

For this reason, it is advantageous if the local maximum value is regarded as the deviation of the peak value of the measured value from the average value.

When in this way the permissible bandwidth for fluctuations of the mean value and, on the other hand, for the maximum values thus taken as deviations from the mean value are determined,
Reliable, time-gap-free quality information is extracted from only six measured values read out periodically. This allows the average value to be prepared continuously for periodic reading out in a very simple manner. This is because the actual measured values that occur are prepared via a low filter.

Up to now, according to the invention, the measured values were stored as analog signals and provided for periodic readout. However, it is also possible to prepare the measured values in digital form, which provides advantages for subsequent processing.

This is done as follows. i.e. dividing each measuring field into measuring steps and generating a continuous output signal for each measuring step reached or exceeded by the successively occurring measured value of the measuring point; are read out and detected by a central data detection device at predetermined time intervals.

It is true that this method does not allow the time course of the measurand to be detected. However, it is detected which measured quantity occurred during the readout time interval. As a result, local maxima of the measured parameter occurring during the readout time interval are detected. This provides information on the course of the measured values during the readout time interval. By storing only the maximum value and detecting the maximum value when reading, the amount of data to be transmitted can be reduced.

The output signal is preferably read out by the central data processing unit and erased again after detection. This allows the measurement data course and the quality of the process to be elucidated by evaluating the measurement data of several consecutive readout time intervals. The length of the readout time interval can be chosen to be very short. The reason is that the measurement data has already been digitized. The shortening of the readout time interval makes local maximum monitoring and evaluation essentially equivalent in time to continuous monitoring and evaluation. Nevertheless, the readout (scanning) frequency is below the frequency required by conventional methods of gap-free measurement data detection.

Embodiments Next, the present invention will be explained in detail with reference to the drawings, with reference to the illustrated embodiments.

FIG. 1 shows a temporally continuous measurement record of measured values, for example the thread tension of a thread 1 running in a textile machine. A portion of this type of textile machine is illustrated in FIG.

FIG. 6 schematically shows four identical processing stations of a large textile machine with yarn 1. In FIG. Each thread 1 is Godets 2
is introduced into the stretching region, where it is guided through a heating device 3 and is removed again via a cordet 4. 5
indicates a yarn pulling tension (yarn tension) measurement value generator. In the winding device, each thread is wound onto a bobbin 7 via a drive roller 6. Each measured value generator 5 includes a measuring sensor 8 and a comparison circuit 9, which are illustrated in detail in FIG.
Consists of. These will be described later.

The measured yarn tension force, illustrated in FIG. 1, remains within a predetermined range in the absence of process disturbances. Ideally, the measured value is constant. However, fluctuations, especially short fluctuations in time, also occur.

Analog measurement value signals of this type can only be detected without gaps in time if the readout is carried out consecutively in time. However, especially when there are a large number of measuring points, as occurs in machines for processing or generating large amounts of yarn in the textile industry, time-free detection and central processing only require large computer capacities. This requires almost unrealizable costs in terms of cables and circuitry. As a rule, the measurement locations are therefore read out one after the other. Since a certain amount of time is required for each readout, the following results from such sequential readouts: In other words, it is not possible to detect the measured values without gaps in time, and rather a minimum time is required for reading out the measured values, which is determined not only by the necessary AD conversion but also by the necessary AD conversion. In order to obtain a sufficiently high resolution of the measured values with a small scanning time, the number of measuring points connected to the computer must be limited.

However, even this does not detect short-term fluctuations, which may be of industrial significance.

According to the invention, first of all a predetermined bandwidth of the measurement values that occur or are to be detected is determined.

This bandwidth is illustrated in FIG. 1 by a dashed line. This bandwidth of measurements is defined in the present invention as the measurement region.

In the exemplary embodiment, the bandwidth of this measurement is divided into measurement steps 1. each having the same size. It is divided into II and III■.

Next, based on FIG. 2, how can such measured values that occur without any gaps in time be transformed according to the present invention?
These measured values are determined at a predetermined interval ST1
．． ST2. The manner in which it is prepared for reading in ST3 will be explained.

According to the invention, a predetermined digital output signal AI to A■ corresponds to each measurement area I to ■. The output signals AI to A■ are called up when or as soon as the actual measured value passes through the corresponding measuring stage. Furthermore, the output signals AI to A2 are stored as long as they are recalled. The called output signal is therefore always ready for reading. A total of output signals All to A2 are produced in the time interval 1.1. These output signals are stored and therefore ready for reading at the reading time STI. It is clear from this that very large fluctuations in the measured values occurred during the readout time interval. The fluctuation range in the readout time interval T2 is larger, while much smaller fluctuations occur in the readout time interval T0.

By the way, the time intervals TI, T2. T3. The values stored at time ST1 . It can be called up and subsequently deleted by the central computer in ST2. Therefore, a temporally seamless detection of the course of the measured values does not take place.

However, the maximum value can be detected without any time gap. In particular, the readout times TI, T2 . By shortening T3°, it is possible to almost completely elucidate the measured value course, for example with regard to periodic measured value fluctuations, the time course of the measured value dispersion, the time trend of maximum values, etc. In this case as well, the readout times TI, T2. T is significantly larger than the readout time of conventional methods, since the local maxima of the measurements occurring in the analogue can be detected reliably. Moreover, only randomly occurring measurement data are read out at the time of readout (that is, the readout can be carried out at defined time intervals and at defined points in time, so that especially extreme measured values and measured value peaks are randomly generated). Moreover, the disadvantage of having to provide costly intermediate data processing devices that cannot be detected well is avoided.

In order to divide the measurement area into measurement stages, a comparator circuit (fourth
fig.) is provided as a particularly advantageous embodiment. When the comparison circuit associates each measured value with a comparison value in a predetermined series of steps, and the respective steps of the measured value and the comparison signal satisfy the criteria for comparison, e.g. are equal,
An output signal is generated that corresponds to the respective measurement step.

Each comparison circuit consists of eight comparators 1o, each of which has an input 11 connected to a series-connected resistor 12.

The comparator outputs an output signal A1 or Av as soon as the value at input 13 reaches the voltage value at input 11.
It is configured to send out I. Each resistance 1
2 are respectively defined to represent a measuring stage, the respective output signals AI to A■ correspond to measuring stages I to ■, respectively.

The circuit can also be configured so that the measurement steps overlap slightly. In this case, for example, mutual locking means for the measuring stages can prevent the signals of the two measuring stages from being stored for measurement values in the overlapping region of the two stages. In any case, only the highest value is sub-stored. For this purpose, the comparator is connected to a logic switch circuit, not shown in detail, which eliminates the output signals AI to A2 as soon as the next higher signal is called.

As shown in detail in FIGS. 6 and 4, each comparator circuit has a respective output value AI to A
is connected to a memory device 14 in which is stored. Output signals AI to A2 are present in the memory and are supplied to the parallel-to-serial converter 15 when they are recalled during the readout period. The function of the parallel-to-serial converter is to convert the parallel and simultaneously generated signals AI to A■ into a pulse train. The pulse train can be fed to the computer 17 via only one line 16. The memory device 14 is also provided with the previously described logic switch circuit which erases the output signals AI to A■ when the next higher output signal is generated.

The readout time is predetermined by the computer 17 (FIG. 6). At the time of readout, the computer in each case sends out a coded signal sequence via a connected parallel-to-serial converter 18, if necessary for addressing via a line 19. As a result, one memory 140 responds in each case and the stored output signals AI to A2 of the measurement location are read out. When a data call is made via line 16, the computer sends out an erasure signal via line 19, which is encoded in the individual memory, so that the memorized output signal of this measuring point is erased. and empty for the next read time interval.

Furthermore, rather than evaluating the entire detection range of the measured value generator and likewise the entire range of possible measured values, it is advantageous to select from the possible measured values a measured value bandwidth that represents normal operation. be. Due to the limitation of the measuring range detected by the invention, measured values lying outside the working range are fundamentally excluded from the evaluation and thus can be used without any restriction on display accuracy (further simplification). If the technical means are the same, display precision within one selected measurement area can be sharpened.

FIG. 5 illustrates the measuring method according to the invention in preparing the analog output signal at each measuring point. The abscissa is the time axis common to all diagrams. Readout time intervals 81 to S4 are illustrated on the time axis. The ordinate of the diagram I shows, for example, a possible course of the measuring voltage U of the measuring sensor. The measured voltage represents an actual, continuously detected measurement value. The diagram ■ shows the average value Umxtte□ of the measured voltage in the ordinate. According to the invention, this average value is generated by passing the measured values through a low-pass filter.

The diagram ■ shows the formation, preparation (holding) and reading out of the maximum value U ax in the course of its time.

Diagram 3 likewise shows the time course, formation, preparation (holding) and readout of the minimum value Umln of the measured voltage.

The measured voltages are shown in diagrams ■ and ■, respectively.
Shown in relation to a reference voltage (average value). For this purpose, in particular, a voltage zero or an average value is taken into account. Local maximum value formation and preparation (holding) will be described in detail below.

For example, in the readout time interval S3, the measured voltage is initially essentially constant in the time domain a and equal to the average value. Therefore, the maximum values UmaX and Umln also correspond to the average value in this time interval. However, in the next time range, as diagram I shows, a sharp and short-lived increase in the measured voltage U occurs. Due to the short-time nature of this maximum value, this variation does not show any influence on the average value. However, the maximum value Umax follows the increase of the measured value up to a maximum value and then takes this maximum value. This maximum value then coincides with ζ during the readout time interval S6 and is centered at the end of this readout time interval, as shown in the diagram ■.

is read out by the data detection device and then erased again, so that the value provided at the measurement location returns to the reference value again.

Following the instant in the read-out time interval S6, some deviations in the negative direction from the average value, in particular a sharp short drop in the measured voltage at KC, are shown. As shown in the diagram ■, the minimum value memory Um□. is also the actual measured value U,
Depending on this negative deviation from the average value, the lowest value in each case is then stored and retained for reading at the end of the reading time interval S3.

At the end of the readout time interval $3, the value of the mean value and the maximum positive and negative deviations therefrom are prepared or retained, indicating the course of the readout time interval, for readout and evaluation by the central data detection device. Ru.

For this purpose, in addition to the average value, an absolute maximum value can also be prepared or maintained.

In this way, for example, the yarn tension in a textile machine having multiple processing points is continuously detected for each yarn to be processed, the maximum, minimum and average values are prepared for each readout time interval, and the The data can be read out at time intervals from processing location to processing location via a scanner by means of a data acquisition device and evaluated in order to represent quality information. At that time, the read value Um is □. A fault alarm signal can be generated depending on □, Umax and Umln. For this purpose, a predetermined bandwidth is first determined in advance for the average thread tension, which is indicated by the output signal Umkte□. If the actually determined average value leaves this bandwidth of the permissible thread tension, a fault warning signal is emitted and, for example, the processing station in question is stopped or a fault is signaled to this processing station. Similarly, a predetermined bandwidth is predetermined for the maximum deviation from the average value. A predetermined bandwidth is likewise predetermined for the minimum value as a deviation from the average value. This bandwidth can then have different dimensions.

FIG. 6 is a basic circuit diagram for preparing analog maximum values and analog average values at each measurement location. The measured value, for example the thread tension at a measuring point as shown in FIG. 6, is continuously detected by a detector or a measured value generator 5. Reference numeral 14 designates a memory device corresponding to each measurement location. In this memory device, the measured values are amplified and processed into a maximum value UmaX, an average value Umlt and a minimum value Un1, nK and stored for readout. For this purpose, the memory device 14 first has an amplifier 20 for amplifying the measurement signal. Then, a maximum value is formed as a deviation from the average value in a peak value measuring circuit 21, and an inverted peak value measuring circuit 22
The formation of a minimum value as a deviation from the average value takes place in . Reference numeral 23 indicates a low-pass filter for generating the average value. Other electronics of the memory device are not shown. At the output of the memory device, a maximum value of the measured value per readout time interval, a continuous average value and a maximum value of the measured value per readout time interval are provided accordingly.

These values are fed to a switching device, a so-called "scanner" 31, whose function is to feed parallel and simultaneous signals sequentially via an AD converter 32 and a line 16 to a computer 17.

As already mentioned, the reading time is predetermined by the computer 17. At each readout time, the computer sends out a coded signal sequence for addressing via line 19, if appropriate.

As a result, the measuring point with the corresponding memory in each case reacts and the stored output signal is read out. After a data call has been made via line 16, the computer sends out via line 19 an erasure signal which is encoded on the basis of the individual memory, so that the stored maximum value or minimum value is will be deleted. The average value is maintained. Furthermore, a fault alarm signal device 24 can also be activated via line 19.

The peak value measuring circuit and low-pass filter used in this circuit are well known. The peak value measuring circuit will be explained once again based on the simplified circuit diagram shown in FIG. The peak value measuring circuit for calculating the maximum value is particularly composed of a diode 25 and a capacitor 2.
It has 6. The capacitor is connected to a voltage of Ov. Voltages of 0 to 10 V are permissible for the measured value U. Since the diode 25 is polarized in the direction of current interruption, the capacitor 26 is charged to the maximum value reached at the time of readout, so that this maximum value is stored and maintained as the output signal for the readout time interval.

27 is a switch that short-circuits the capacitor 26. Switch 27 is controlled via net 19 when the maximum value Ue can be erased after reading.

The inverted peak value forming circuit 22 has a diode P28 and a capacitor 29. However, the polarity of the diode is opposite to that of diode 25. The capacitor, on the other hand, is connected to a voltage of 15 V, which is higher than the maximum possible measured voltage, previously assumed to be 10. That is,
The respective drop in the measured value on the capacitor 29 appears as an increasing and persistent voltage gradient, which is retained at the output of the memory as a minimum value Umm. 30
Shown is a switch capable of shorting the capacitor and thus sending out the erase signal via line 19.

The low-pass filter 23 consists essentially of resistors and capacitors in a well-known manner.

The circuit of FIG. 7 differs from the circuit shown in FIG. 6 in the following respects. That is, in FIG. 7, local maximum values are displayed by their absolute values. Similar to FIG. 6, the measured values can also be processed in a memory device such that the local maximum value is represented as the difference between the measured local maximum value and its average value. There are two alternative possibilities for this.

On the one hand, means are inserted between the circuit point 38 and the voltage follower 33 and between the same circuit point 38 and the voltage follower 34 to form a difference between the actual measured value and the average value U. be able to. Such a differential amplifier is shown at 39 in FIG.

Another possibility is that the voltage followers 35 and 3 of FIG.
6, a differential amplifier connected to the output side of the differential amplifier 37 is provided. Thereby, the maximum values Umax and Umln are processed to be expressed as the difference between the absolute maximum value and the average value.

Effects of the Invention According to the present invention, the maximum and minimum values within a time interval are memorized and read out and detected by the central detection device at predetermined time intervals. There is an advantage in that information can be obtained in a timely manner compared to the process that should be carried out.

[Brief explanation of drawings]

FIG. 1 is a graph showing an example of a measurement record with no gaps in time of measured values of yarn tension, and FIG. 2 shows how the measured values of FIG. 1 are measured using the method of the present invention. FIG. 6 is a block diagram schematically showing a part of a textile machine having four measuring points and the associated electrical equipment; FIG. 5 shows a detailed circuit diagram of the comparator circuit of FIG. 5 together with a series converter; FIG. 5 is a voltage waveform diagram illustrating the use of the method of the invention in preparing an analog output signal at each measurement point; FIG. , FIG. 6 is a block diagram of a circuit for preparing analog maximum values and analog average values at each measurement location, FIG. 7 is a circuit diagram of a simplified peak value measurement circuit, and FIG. 8 is a circuit diagram of an alternative embodiment of the circuit of FIG. 7, additionally provided with a differential amplifier; FIG. 5 Measured value generator, 8 Measurement sensor, 9 Comparison circuit, 14 Memory device, 17 Calculator, 21 Maximum value formation circuit, 22 Minimum value formation circuit, 23 Average value formation circuit, Continuing from page 1 Priority claim ■May 30, 19 [phase] Nishi Toy 01 Kenw hand August 17 0 Nishi Toy 0 Inventor Karl Werner Frörichts (D E) @P3420163.7 Tsu ( DE) [Phase] P3430223.9 Federal Republic of Germany Wuppertal Lulangerfeld in Der Fist Freude 121

Claims (1)

[Claims] 1. A method for centrally detecting measured values at a plurality of measuring points when data generated at each measuring point is repeatedly transmitted at time intervals to a central data detection and/or data processing device. , the maximum and minimum values of successive measured values within each time interval are detected and stored each time as an output signal, and the stored output signals are transferred to the central data at predetermined short time intervals. A method for centrally detecting measured values at a plurality of measuring points, characterized in that the measured values are read out and detected by a detecting device. 2. A method for centrally detecting measured values at a plurality of measuring points as claimed in claim 1, wherein the stored output signals are read out and erased by a central data processing device after detection. 2. Measurement values of a plurality of measurement points according to claim 1 or 2, in which a limited bandwidth of measurement values is selected as the measurement area from possible measurement values or detection areas of the measurement value generator. Central detection method. 4. A method for centrally detecting measured values at a plurality of measuring points according to claim 6, in which a value exceeding a selected measurement area is signaled as a fault. 5. A method for centrally detecting measured values at a plurality of measuring points according to any one of claims 1 to 4, wherein an average value is continuously formed from the measured values, stored, and read out. 6. The local maximum value is formed by comparison with the average value and a fault signal is emitted when the local maximum value thus obtained exceeds a predetermined maximum permissible deviation. The method for centrally detecting measured values at a plurality of measurement points according to item 5. Z. A method for centrally detecting measured values at a plurality of measuring points according to claim 5, wherein the maximum value occurring within a readout time interval is detected as the maximum deviation of the measured value from the average value, stored and read out. 8. Central detection of measured values at a plurality of measurement points according to any one of claims 5 to 7, wherein the average value is detected from continuously actually measured values, regardless of the readout time interval. Method. 9. A method for central detection of measured values of a plurality of measuring points according to claim 8, wherein the average value is formed by feeding successively occurring actual measured values to a low-pass filter. Wave 10 Any one of claims 1 to 9 in which the local maximum value is detected by measuring the high value of the actual measured value.
Central detection method for measured values at multiple measurement points as described in Section 2. 11. Divide the entire measurement area of each measurement point into measurement steps (from 1 to 2), and measure each measurement step at which the measured value reached or exceeded at each measurement point. A plurality of i!s according to any one of claims 1 to 10 for generating an output signal (AI to l) and storing and preparing it for reading. 111 Central detection method for measured values at fixed locations. 12. feeding a stepwise increasing reference value in each case to a comparator and comparing it with the measured value; each comparator detects when the measured value at least approximately reaches or exceeds the reference value; 12. A method for centrally detecting measured values at a plurality of measuring points as claimed in claim 11, wherein an output signal is generated.