Method and device for highly accurate determination of the level of a product in a container

The invention relates to a method for highly accurate determination of the filling level of a filling material in a container, wherein measuring signals are emitted in the direction of the surface of the medium and reflected on the surface as echo signals and wherein by evaluating the amplitude and phase values ​​of the reflected echo signals of the level is determined in the container via a pulse transit time method. Can be applied both the pulse radar method and the FMCW method, in which continuous waves periodically linear, z. B. with a sawtooth voltage, are frequency modulated. Furthermore, the invention relates to a device for highly accurate determination of the level of a product in a container.

Pulse transit time methods exploit the physical law, according to which the travel distance is equal to the product of transit time and propagation speed. In the case of level measurement, the running distance is twice the distance between the antenna and the surface of the medium. The actual useful echo signal and its duration are based on the so-called. Echo function or determines the digital envelope, the envelope represents the amplitudes of the echo signals as a function of the distance 'antenna - surface of the medium'. The level itself is then obtained from the difference between the known distance of the antenna from the bottom of the container and the distance determined by the measurement of the surface of the medium from the antenna.

DE 31 07 444 A1 describes a high-resolution pulse radar method. A generator generates first microwave pulses and emits them via an antenna at a predetermined transmission repetition rate in the direction of the surface of the medium. Another generator generates reference microwave pulses that are the same as the first microwave pulses, but differ slightly in terms of repetition rate. The echo signal and the reference signal are mixed. At the output of the mixer is an intermediate frequency signal. The intermediate frequency signal has the same shape as the echo signal but is stretched by a time expansion factor equal to a quotient of the transmission repetition rate and the frequency difference of first microwave pulses and reference microwave pulses. At a transmission repetition rate of a few megahertz, a frequency difference of a few hertz and a microwave frequency of a few gigahertz, the frequency of the intermediate frequency signal is below 100 kHz. The advantage of the transformation to the intermediate frequency is that relatively slow and therefore cost-effective electronic components for signal detection and / or signal evaluation can be used.

The signal evaluation takes place via the so-called Envelope curve. The envelope itself is the result of a rectification, optionally a logarithmization and a digitization of the intermediate frequency. The determination of the distance takes place via the determination of the distance of the wanted echo signal, which represents the fill level, to a reference signal. The number of sampling points between the two maxima at a constant sampling time is directly proportional to the distance 'antenna - surface of the medium'.

In order to increase the accuracy of measurement, it has already become known, in addition to the maxima (peaks), the amplitude information provide, and their phase positions for evaluation use. For this purpose, the amplitude-modulated intermediate frequency is demodulated and decomposed into its complex components. This is achieved z. B. by the so-called Quadrature demodulation, d. H. the intermediate frequency is multiplied once by a sine wave (Q) and once by a cosine wave (I), both oscillations having a similar frequency to the intermediate frequency. The high frequencies resulting from the multiplication are filtered out with a Taßpaßfilter. The envelope signal HK is obtained from the root of the sum of the squares of I (in-phase component) and Q (quadrature component): HK = <maths id="math0001" num=""><math display="inline"><mrow><msqrt><msup><mrow><mtext mathvariant="italic">I</mtext></mrow><mrow><mtext>2</mtext></mrow></msup><mtext> + </mtext><msup><mrow><mtext mathvariant="italic">Q</mtext></mrow><mrow><mtext>2</mtext></mrow></msup></msqrt></mrow></math><img file="EP1069438A1_D0001.tif" /></maths>, Subsequently, the usual amplitude evaluation takes place; At the found locations of the maxima, the respective phase position and the difference of the two phase positions are additionally determined. The distance of the antenna from the surface of the medium is then composed of a proportion of integer wavelengths, which results from the amplitude evaluation, and a phase remainder.

The known methods provide reliable results as long as it is ensured that the signals received by the antenna are reflected exclusively on the surface of the medium (incidence and reflection take place in the direction of the normal). Once the so-called. Multipath propagation occurs, the measurement accuracy deteriorates. Multipath propagation means that the echo signals not only contain the actual useful signal, but in addition a Störsignalanteil, which is due to retroreflections of the measuring signals on the vessel wall or other located in the container interior fittings.

Previous proposals for solving this problem are limited to the occurrence of interference from the outset exclude. According to a first embodiment, unwanted retroreflections can be prevented by positioning the antenna so far away from the container wall (or other interfering retroreflector) that the impact of measuring signals on these 'interfering retroreflectors' can be reliably excluded.

The disadvantages of this circumvention are obvious; especially with the considerable dimensions of the storage container as z. B. used in petrochemical, the subsequent installation, repair or replacement of the level in the central region of the container lid is much more difficult to accomplish than in the edge region. In addition, there is another source of error in the measurement of the level, which beats the more serious book, the farther the meter is removed from the edge of the container lid: Usually, the lid of large storage containers, such as storage tanks or silos, an outwardly curved or tapered Shape. If the side walls of the container bulge out, which can be consistently observed in the region of the maximum filling level of the container, then the lowering of the lid and thus the change of the reference distance 'antenna - container bottom' in the middle region has a greater effect than in the edge region.

Furthermore, attempts have been made so far to get the problem of multipath propagation under control by using an antenna with optimized directional characteristics. Of course, such antennas tuned to the individual case are very expensive, which increases the cost of the level measuring device.

In certain applications, it is necessary to replace the free field measurement for determining the level by a measurement using a stillpipe. Schwallrohre be z. B. always used when the measurement result would be falsified by corrugation of the contents within the container.

As previously mentioned, the level over the term of the electromagnetic waves at twice the distance 'antenna - surface of the medium' is determined. Thus, the measurement result is correct, therefore, the propagation speed of the electromagnetic waves in the space between the antenna and contents must be known exactly. Approximately, the propagation velocity in a large number of applications is equated with the propagation velocity in air. However, this assumption is only valid if the transverse dimensions of the space in which the electromagnetic waves propagate are large compared to the wavelength.

If the measurement of the fill level is made via a stillpipe, the above assumption is only approximately correct. A small diameter tube or other container over which the waves propagate acts as a waveguide. Since the propagation of electromagnetic waves with the speed of light applies only in the ideal case, that wave packets can propagate undisturbed in free space, dispersion occurs in waveguides - the propagation velocity is thus dependent on the wavelength. The decisive for the propagation of a wave packet group velocity in a waveguide is smaller than the speed of light.

If the wavelength is in the order of magnitude of the propagation space, there are several modes capable of propagation which, due to the dispersion, have different propagation velocities (→ multimode propagation). The consequence of the multimode propagation is the occurrence of constructive and destructive interference between the measurement signals, which in turn leads to an oscillating measurement error being superimposed on the actual fill level measurement signal within a critical measurement range and corrupting it. Due to the oscillation, the contradictory case may occur that z. B. despite actual level decrease of the measuring apparatus, an increasing level value is displayed. Of course, such an obvious measurement error is completely unacceptable.

Incidentally, the critical measuring range lies between the zero running distance and a maximum running distance at which the measuring error disappears, since the wave packets have traveled so far apart that interference no longer occurs. This maximum running distance depends on the nature of the waveguide and the frequency of the electromagnetic waves used.

Again, this problem can be curbed by the fact that very high quality waveguides are used. In addition, the antenna must be optimally adapted to the respective diameter of the waveguide. However, a cost-effective solution is characterized precisely by the fact that the tubes do not have to be of high quality and that it is also possible to use a low-cost antenna which is able to cover any inner diameter of surge tubes acting as a hollow conductor.

The invention has for its object to provide a method and an apparatus that allow highly accurate level measurements in containers - regardless of the occurrence of multipath propagation or of multi-mode propagation.

The invention is achieved with respect to the method in that errors that occur in the determination of the level and caused by noise that constructively or destructively interfere with the regularly reflected on the surface of the medium echo signals are compensated. Cause of errors that occur as a result of multipath propagation, are constructive or destructive interference between the actual useful echo signal, which is reflected at the surface of the medium, and the proportion of the useful echo signal, which from a retroreflector, z. B. the container wall, is reflected. If x 'is the distance traveled by the actual useful echo signal which has been reflected at the surface of the medium, and x "is the longer path of the useful echo signal which has been reflected by another retroreflector, in particular on the container wall, interferences occur if the path difference satisfies the condition Δx = n · λ / 2, where n is an arbitrary integer. The maxima of the two signals are so close together that the resolution of the electronics is not sufficient to separate the two peaks. By means of the method according to the invention, these errors can now be reliably eliminated, which, expressed in numbers, means that the error tolerances in the area of ​​fill level measurement are reduced by one order of magnitude compared with the previously known methods. The same problem occurs, as described above, also as a consequence of multimode propagation of wave packets in stillpipes or other containers carrying the wave packets. The method according to the invention is therefore also ideally suited to eliminate measurement errors due to multimode propagation.

The advantages of the solution according to the invention can be seen in the fact that the antenna can also be positioned in the edge region of the container or in the area of ​​another retroreflector located in the interior of the container without the interference signals distorting the measurement results. The previously required minimum distance from a Störreflektor can now be easily overcome. Since it is possible to selectively eliminate the portion of the interfering signals from the wanted echo signal, an antenna can be used in addition, although optimally adapted to the reflection properties of the respective filling, but no optimized directional characteristic as a function of the respective desired mounting position must have.

In the case of multimode propagation, the advantage is also significant cost savings, since on the one hand no high-quality waveguide must be used; On the other hand, an antenna for a plurality of waveguides or surge tubes can be used with different dimensions.

According to an advantageous development of the method according to the invention, it is provided that a correction value is determined and stored for selected fill level values ​​at least within a critical fill level measuring range in which the interference signals are generated.

A preferred embodiment of the method provides in particular that the start level is measured; then the correction value is initialized to zero; starting from this start level, at least the critical Füllstandsmeßbereich z. B. successive and stepwise through; then a correction value is calculated and stored for each selected fill level value within the fill level measurement range.

In order to reduce the number of actually measured within the critical Füllstandsmeßbereichs values ​​to a minimum, proposes an advantageous embodiment of the method according to the invention that for a complete provision of the correction values, the correction values ​​between two z. B. consecutive level values ​​are each interpolated.

A particularly advantageous embodiment of the method according to the invention provides that the correction values ​​for successive fill level values <i>q</i>_<i>K</i><i> , q</i>_<i>K</i>₊₁ be determined as follows: <i>q</i>_<i>K</i><i> = q</i>_<i>O</i> the start level. At start-up level, the correction value is given a predetermined value, e.g. Initialized with zero; Subsequently, the changed level is changed<i>q</i>_{<i>K +</i>}₁ the difference of the corrected amplitude removal values ​​A '(<i>q</i>_<i>k</i>) and the phase removal value. Thus, a new correction value is determined and stored for each subsequently approached level. The calculation of the new correction value preferably takes place according to the following formula:<maths id="math0002" num=""><math display="block"><mrow><mtext>K (</mtext><msub><mrow><mtext mathvariant="italic">q</mtext></mrow><mrow><mtext mathvariant="italic">K</mtext><mtext>+1</mtext></mrow></msub><mtext>) = φ (</mtext><msub><mrow><mtext mathvariant="italic">q</mtext></mrow><mrow><mtext mathvariant="italic">K</mtext><mtext>+1</mtext></mrow></msub><mtext>) - A '(</mtext><msub><mrow><mtext mathvariant="italic">q</mtext></mrow><mrow><mtext mathvariant="italic">K</mtext><mtext>+1</mtext></mrow></msub><mtext>) = φ (</mtext><msub><mrow><mtext mathvariant="italic">q</mtext></mrow><mrow><mtext mathvariant="italic">K</mtext><mtext>+1</mtext></mrow></msub><mtext>) - A (</mtext><msub><mrow><mtext mathvariant="italic">q</mtext></mrow><mrow><mtext mathvariant="italic">K</mtext><mtext>+1</mtext></mrow></msub><mtext>) - K (</mtext><msub><mrow><mtext mathvariant="italic">q</mtext></mrow><mrow><mtext mathvariant="italic">K</mtext><mtext>+1</mtext></mrow></msub><mtext>).</mtext></mrow></math><img file="EP1069438A1_D0002.tif" /></maths>

The embodiment of the invention exploits the fact that the phase evaluation is much more sensitive in large areas and thus provides more accurate measurement results than the amplitude evaluation. If relatively large measurement errors already occur in the amplitude evaluation as a result of multipath propagation or multimode propagation, the measurement errors that the phase evaluation delivers are still relatively small. Only when the measurement error in the amplitude evaluation is greater than λ / 4 (which corresponds to a measurement error of 12 mm, for example, at a frequency of 6.3 GHz) occurs in the phase evaluation, a phase jump, which causes a measurement error of 24 mm. The error from the phase evaluation is in the range of a phase jump thus twice as large as the error from the amplitude evaluation. Since it provides the inventive method to avoid such phase jumps, one achieves the desired high measurement accuracy of approx. 1 mm in the determination of the filling level.

Is each level within the measuring range, in the multipath propagation or Multimode propagation is to be feared associated with a correction value, it is subsequently possible to correct measured filling level values ​​accordingly. No problems now prepares the hitherto critical case that the level gauge is turned off for some reason. If the correction values ​​in the critical fill level measuring range are known without gaps, after the meter has been switched on, that correction value is used to correct the currently measured fill level, which has previously been determined and stored for the corresponding fill level.

Nevertheless, it may happen that the meter is turned off at a time when the correction values ​​are not yet present over the entire critical Füllstandsmeßbereich. Here are two cases to distinguish: If it is ensured that the level has not changed during the off period of the meter, so provides an advantageous embodiment of the method according to the invention that before the shutdown last stored correction value is used for initialization.

In the event that the correction values ​​in the critical Füllstandsmeßbereich are not fully known and, moreover, if it can not be ruled out that the level has changed during the off period, then the correction value is initialized to zero after a shutdown of the level measuring; For each correction value, a status value is additionally stored, which provides information about the reliability of the respective correction value. These status values ​​can be, for example, 'safe', 'uncertain', 'indefinite'. In the course of the further measurements, the determined correction values ​​are adapted to the already stored correction values, whereby the newly determined correction values ​​are optionally corrected.

A variant of the method according to the invention provides that an unambiguous phase correction value (P (q)) is determined on the basis of the determined correction value, where P = n · λ / 2 with n∈ Z, and that the phase removal value is corrected in accordance with the unique correction value. Furthermore, it is proposed that gray areas are defined on which no clear phase correction value is defined after a switch-off process and a small correction value (K (q)).

The invention will be explained in more detail with reference to the following drawings. It shows:<ul id="ul0001" list-style="none"><li>1 shows a schematic representation of a first embodiment of the device according to the invention,</li><li>2 shows a schematic representation of a second embodiment of the device according to the invention,</li><li>3 is a diagram showing the measurement errors from the amplitude evaluation and the phase evaluation at different levels,</li><li>4 shows a flow chart for determining the correction values ​​according to a first embodiment,</li><li>5 shows a flowchart for determining the correction values ​​according to a second embodiment,</li><li>6 shows a schematic representation of the stored correction values ​​K (q) as a function of the fill level (q) and</li><li>7 shows a schematic illustration of the correction values ​​after a shutdown of the fill level measuring device.</li></ul> FIG. 1 shows a schematic representation of a first embodiment of the device according to the invention. A product 4 is stored in a container 2. For the determination of the level F is the level measuring device 1, which is mounted in an opening 7 in the edge region of the lid 6. Via the antenna 8, transmission signals generated in the transmission circuit 9, in particular microwaves, are radiated in the direction of the surface 5 of the filling material 4. At the surface 5, the transmission signals are partially reflected. The reflected echo signals are received and evaluated in the reception / evaluation circuit 10. The correct timing of transmission of the transmission signals and reception of the echo signals via the transmitting-receiving pair 12th

If the current level q is within a critical level measurement range <i>F</i>_{<i>critical</i>}Thus, multipath propagation occurs, i. h., The received echo signals in addition to the regularly reflected on the surface 5 of the medium 4 signals also interference signals that have not been reflected on the surface 5 of the medium 4, but on the container wall 3 or other located in the container 2 internals. The internals may be, for example, a stirrer. If the path difference Δx = x'-x "between the regularly reflected echo signals and the interference signals is an integer multiple of ± λ / 2, constructive and destructive interferences occur. As a result of the interference fill level echo signals appear in the envelope, which are so close to each other that they are no longer resolved by the evaluation circuit commonly used. The echo signals are shown in FIG. 1 also shown schematically as a function of the running distance (in sampling points). By means of the method or the device according to the invention measuring errors due to amplitude fluctuations are eliminated.

While Fig. 1 shows the measurement in the free field, Fig. 2 a schematic representation of a second embodiment of the device according to the invention, in which the measurement is carried out via a surge pipe 14. The stillpipe behaves - physically seen - like a waveguide. The stillpipe 14 is guided through an opening 15 of a floating roof 13 and extends into the vicinity of the bottom of the container second Incidentally, a floating roof 13 is preferably used when the formation of an explosive gas space above the contents 4 is to be avoided. While in the case of free-field propagation of the electromagnetic waves, the multipath propagation is responsible for the spurious signals in the reflected level echo signal, in the case of the stillpipe measurement, the multimode propagation causes the spurious signals. The influence of both interference signals on the quality of the Füllstandsmeßwertes can be eliminated by means of the inventive solution.

FIG. 3 shows a diagram which represents the measurement errors resulting from the amplitude evaluation and the phase evaluation, depending on the respective level. The solid line indicates the measurement errors which have been determined on the basis of the amplitude evaluation; the dashed line indicates the measurement errors from the corresponding phase evaluation. It can be clearly seen that the phase errors are much less sensitive to multipath or Multimode propagation are the corresponding amplitude measurement errors. Serious measurement errors occur in the phase evaluation only when the measurement error resulting from the amplitude evaluation, exceeds a certain limit. This limit corresponds to a value of ± λ / 4. At a wavelength of the measurement signals of 6.3 GHz, this limit corresponds to a measurement error of 12 mm when determining the fill level. Up to this limit, a very high accuracy is achieved by evaluating the phases of the echo signals. In order to avoid the sudden increase in the measurement error in the phase evaluation, the advantageous refinement of the method according to the invention described below with reference to a flow chart is used.

In Fig. 4 is a flowchart for determining the correction values ​​K (q) according to a first embodiment indicated. These correction values ​​K (q) make it possible to correct a fill level value measured in a container 2 in such a way that noise signals which correspond to the actual fill level echo signal due to multipath propagation or Superimpose multimode propagation, be eliminated. In principle, relatively large measurement errors are to be expected if the resolution of the Füllstandsmeßgeräts is not sufficient to accurately determine the position of two closely spaced maxima of the amplitude evaluation. This is - as already described above - the case as soon as multipath propagation in free field or Multimode propagation in the stillpipe occurs: Interference interferes constructively and destructively with the actual level echo signal. Instead of a peak, which represents the level, appear two peaks, resulting in insufficient resolution of the level in a broadening or reflected in the fluctuation of the amplitude maximum.

The program runs as follows: After the program start at point 20, under the program point 21, the current distance <i>A</i>(<i>q</i>_<i>O</i>) to the level <i>q</i>_<i>O</i> measured. If it is the first measured value recording at all, which is checked under program item 23, then the correction value becomes<i>K</i>_<i>old</i><i> =</i> K (q) with the matching value of the stored table (K (<i>q</i>_<i>O</i>) (program item 24). By the way, this table has been initialized to zero or some other value before the first startup. Therefore, every status is indefinite. Under the program item 25, the corrected distance value from the amplitude evaluation<i>A '</i>(<i>q</i>_<i>O</i>) certainly. This is included in the phase evaluation at point 26. Under program item 27 becomes the level<i>q</i>_<i>O</i> a new correction value <i>K</i>_<i>New</i> (<i>q</i>_<i>O</i>) determined according to the following formula:<maths id="math0003" num=""><math display="block"><mrow><msub><mrow><mtext mathvariant="italic">K</mtext></mrow><mrow><mtext mathvariant="italic">New</mtext></mrow></msub><mtext> (</mtext><msub><mrow><mtext mathvariant="italic">q</mtext></mrow><mrow><mtext mathvariant="italic">O</mtext></mrow></msub><mtext>) = </mtext><mtext mathvariant="italic">φ</mtext><mtext>(</mtext><msub><mrow><mtext mathvariant="italic">q</mtext></mrow><mrow><mtext mathvariant="italic">O</mtext></mrow></msub><mtext>) - A '(</mtext><msub><mrow><mtext mathvariant="italic">q</mtext></mrow><mrow><mtext mathvariant="italic">O</mtext></mrow></msub><mtext>) </mtext><mtext mathvariant="italic">= φ</mtext><mtext>(</mtext><msub><mrow><mtext mathvariant="italic">q</mtext></mrow><mrow><mtext mathvariant="italic">O</mtext></mrow></msub><mtext>) - A (</mtext><msub><mrow><mtext mathvariant="italic">q</mtext></mrow><mrow><mtext mathvariant="italic">O</mtext></mrow></msub><mtext>) - </mtext><msub><mrow><mtext mathvariant="italic">K</mtext></mrow><mrow><mtext mathvariant="italic">old</mtext></mrow></msub><mtext>,</mtext></mrow></math><img file="EP1069438A1_D0003.tif" /></maths>

The new correction value <i>K</i>_<i>New</i> (<i>q</i>_<i>O</i>) is then saved (program item 28). At point 29 becomes<i>K</i>(<i>q</i>_<i>O</i>)<i>= K</i>_<i>New</i> (<i>q</i>_<i>O</i>) set. The determined and stored correction value provides information about how a measured level echo signal under the given conditions in the container 2 or in the stilling tube 13 must be corrected in order to provide highly accurate information about the current level q of the medium 4 in the container 2.

In Fig. 5 Fig. 10 is a flow chart for determining the correction values ​​K (q) according to a second embodiment. In this case, too, the correction values ​​K (q) allow a fill level value measured in a container 2 to be corrected in such a way that interference signals which correspond to the actual fill level echo signal due to multipath propagation or Superimpose multimode propagation, be eliminated. While in the first case the correction values ​​K (q) are specified in the unit of sampling points or in another distance unit, in the present second case either correction values ​​K (q), the number of corrected wavelengths λ or the phase correction values ​​P (q) are determined and stored.

After the program start at point 30, at point 31, the current distance A (<i>q</i>_<i>O</i>) to the level <i>q</i>_<i>O</i> measured. If it is the first measured value recording at all, which is checked under the program point 32, then the correction value K (q) with the matching value of the stored table K (FIG.<i>q</i>_<i>O</i>) initializes (program item 33). Incidentally, before the first startup, this table has been initialized to zero or some other predetermined value. Therefore, every status is indefinite. Under program item 34, the distance value from the phase evaluation<i>φ</i>(<i>q</i>_<i>O</i>) at the level <i>q</i>_<i>O</i> determined. At point 35 becomes the level<i>q</i>_<i>O</i> a new correction value <i>K</i>_<i>New</i> (<i>q</i>_<i>O</i>) determined according to the following formula:<maths id="math0004" num=""><math display="block"><mrow><msub><mrow><mtext mathvariant="italic">K</mtext></mrow><mrow><mtext mathvariant="italic">New</mtext></mrow></msub><mtext> (</mtext><msub><mrow><mtext mathvariant="italic">q</mtext></mrow><mrow><mtext mathvariant="italic">O</mtext></mrow></msub><mtext>) = φ (</mtext><msub><mrow><mtext mathvariant="italic">q</mtext></mrow><mrow><mtext mathvariant="italic">O</mtext></mrow></msub><mtext>) - A (</mtext><msub><mrow><mtext mathvariant="italic">q</mtext></mrow><mrow><mtext mathvariant="italic">O</mtext></mrow></msub><mtext>) - </mtext><msub><mrow><mtext mathvariant="italic">K</mtext></mrow><mrow><mtext mathvariant="italic">old</mtext></mrow></msub><mtext>,</mtext></mrow></math><img file="EP1069438A1_D0004.tif" /></maths>

The new correction value <i>K</i>_<i>New</i> (<i>q</i>_<i>O</i>) is then stored (program item 36). At point 37 becomes<i>K</i>_<i>old</i> (<i>q</i>_<i>O</i>) <i>= K</i>_<i>New</i> (<i>q</i>_<i>O</i>) set. The determined and stored correction value<i>K</i>_<i>New</i> (<i>q</i>_<i>O</i>) provides information about how a measured level echo signal under the given conditions in the container 2 or in the stilling tube 13 must be corrected to a highly accurate information about the current level <i>q</i>_<i>O</i> of the filling material 4 in the container 2 to deliver. Furthermore, under program item 38, a phase correction value P (<i>q</i>_<i>O</i>). With this, under point 39, the corrected distance value φ '(<i>q</i>_<i>O</i>) calculated according to the following relationship:<maths id="math0005" num=""><math display="block"><mrow><mtext mathvariant="italic">φ '</mtext><mtext>(</mtext><msub><mrow><mtext mathvariant="italic">q</mtext></mrow><mrow><mtext mathvariant="italic">O</mtext></mrow></msub><mtext>) </mtext><mtext mathvariant="italic">= φ</mtext><mtext>(</mtext><msub><mrow><mtext mathvariant="italic">q</mtext></mrow><mrow><mtext mathvariant="italic">O</mtext></mrow></msub><mtext>) - P (</mtext><msub><mrow><mtext mathvariant="italic">q</mtext></mrow><mrow><mtext mathvariant="italic">O</mtext></mrow></msub><mtext>).</mtext></mrow></math><img file="EP1069438A1_D0005.tif" /></maths>

In Fig. 6 is a schematic representation of the stored correction values ​​K (q) as a function of the level q to see. The level q is the value that results from the evaluation of the envelope without correction. The points indicate the fill levels q at which a correction value preferably corresponds to that shown in FIG. 3 shown flowchart has been determined. The measuring points are so close together that the correction values ​​K (q) can be interpolated between the measured fill levels. Each level q subsequently indicated by the meter is corrected by the correction value K (q), thereby eliminating measurement errors caused by multipath propagation or multimode propagation during the measurement process. By means of the method or The device according to the invention thus makes it possible to use a measuring device universally for highly accurate measurement in containers or waveguides with completely different dimensions and geometry.

The critical case already described above is sketched in FIG. 7: The measuring device is switched off at a point in time if not all the correction values ​​K (q) are present within the critical fill-level measuring range. In addition, it can not be ruled out that the level has changed during the switch-off phase.

The correction values ​​in the right part of the curve (solid curve) have the status 'safe'. They were determined and stored, for example, according to the aforementioned method. At a filling level of approx. 16 m, the electricity will be cut off. After switching on the meter, there is no guarantee that the following measurements will be correct. The reason for this is that the correction value K (obtained from the table)<i>q</i>_<i>O</i>) has not yet been determined in a previous measurement cycle. The status is indeterminate, so unknown, and the value of K is any given value, e.g. B. Zero. The course of the curve of the correction values ​​determined after restarting the measuring device is shown in the upper left area of ​​the diagram. The dashed curve shows the course of the correction values ​​K (q) without a power failure. In order to prevent the possibly resulting erroneous measurements, a global status is stored and for each correction value K (q) a status value is stored.

In the illustrated case, the correction values ​​in the right-hand area have the status 'Safe' (solid curve), the correction values ​​K (q) in the upper left area have the status 'Uncertain' (dashed curve) and the correction values ​​in the lower left area have the status 'Safe Corrected' '(dotted curve). Previously not determined correction values ​​have the status 'undetermined' and the predetermined value, preferably zero. Further measurements fuse the individual areas that have a different status. If necessary, the correction values ​​K (q) with different status values ​​must be adapted to each other.

LIST OF REFERENCE NUMBERS

<dl id="dl0001" compact="compact"><dt>1</dt><dd>level meter</dd><dt>2</dt><dd>container</dd><dt>3</dt><dd>container wall</dd><dt>4</dt><dd>filling</dd><dt>5</dt><dd>Surface of the product</dd><dt>6</dt><dd>container lid</dd><dt>7</dt><dd>opening</dd><dt>8th</dt><dd>antenna</dd><dt>9</dt><dd>transmission circuit</dd><dt>10</dt><dd>Reception / evaluation</dd><dt>11</dt><dd>connecting line</dd><dt>12</dt><dd>Transmit-receive switch</dd><dt>13</dt><dd>floating roof</dd><dt>14</dt><dd>stilling</dd><dt>15</dt><dd>Opening in the floating roof</dd></dl><dl id="dl0002" compact="compact"><dt>q, <i>q</i>_<i>O</i></dt><dd>level</dd><dt>K</dt><dd>correction value</dd><dt>φ (q)</dt><dd>Distance value from the phase evaluation</dd><dt>φ '(q)</dt><dd>corrected distance value from the phase evaluation</dd><dt>P</dt><dd>Phase correction value</dd><dt>A (q)</dt><dd>Distance value from the amplitude evaluation</dd><dt>A '(q)</dt><dd>corrected distance value from the amplitude evaluation</dd><dt>S</dt><dd>status value</dd><dt>λ</dt><dd>wavelength</dd></dl>