Method for configuring a magnetic-inductive flowmeter

Abstract

The invention relates to a method for configuring a magnetic-inductive flowmeter (1), wherein the magnetic-inductive flowmeter (1) comprises: a measuring tube (8) for guiding a medium; at least two measuring electrodes (3, 4) and measuring electronics (9), which are electrically connected to the at least two measuring electrodes (3, 4), for determining a flow-velocity-dependent measurement variable; a magnetic-field-generating device (7) for generating a magnetic field penetrating at least sections of the measuring tube (8); comprising the method steps of: determining a first measure of dispersion (e.g. standard deviation) from a flow-velocity-dependent first measurement signal, wherein the first measurement signal is recorded with a first integration time; determining a second measure of dispersion (e.g. standard deviation) from a flow-velocity-dependent second measurement signal, wherein the second measurement signal is recorded with a second integration time, wherein the first integration time differs from the second integration time; determining (e.g. by means of quotient formation) a test value (power b) on the basis of the first and second measures of dispersion.

Description

[0001] Method for configuring a magnetic-inductive flowmeter

[0002] The invention relates to a method for configuring a magnetic-inductive flow meter and a magnetic-inductive flow meter.

[0003] Magnetic-inductive flowmeters are used to determine the flow velocity and / or volumetric flow rate of a free-flowing, electrically conductive medium in a measuring tube. A magnetic-inductive flowmeter always includes a magnetic field-generating device designed to produce a magnetic field perpendicular to the horizontal axis of the measuring tube. This is typically achieved using one or more magnetic coils. To create a largely homogeneous magnetic field, pole pieces can be shaped and attached so that the magnetic field lines run essentially perpendicular to the axis across the entire cross-section of the tube. A pair of measuring electrodes attached to the outer surface of the measuring tube detects an inductively generated electrical voltage that arises when a conductive medium flows along the longitudinal axis of the measuring tube while a magnetic field is applied.Since the measured voltage depends on the velocity of the flowing medium according to Faraday's law of induction, the flow velocity and, with the addition of a known cross-sectional area of ​​the measuring tube, the volumetric flow rate of the medium can be determined from the measured voltage.

[0004] Magnetic-inductive flowmeters measure a flow-velocity-dependent electrical potential between at least two measuring electrodes in a measuring tube through which the medium flows. In addition to the noise resulting from the flow signal, significant disturbances in the process can also contribute to the noise. Sources of noise can be solid particles in the medium that collide with the measuring electrode or turbulence within the medium itself. A typical method for characterizing the noise is frequency analysis of the measurement signal. This usually involves determining and evaluating the power spectral density, i.e., the power of the noise as a function of frequency. The noise is then characterized according to its frequency dependence, with white noise meaning that the power of the noise is independent of the frequency (e.g.,Water with low conductivity), while brown noise means that the power spectral density is proportional to 1 / f. 2is (e.g., pulp and / or paper). Determining the power spectral density requires measuring a long, high-resolution time series of the measurement signal. This procedure typically requires both specialized equipment and expert knowledge. If the exact power spectral density of the noise is known, it is often possible to configure the magnetic-inductive flowmeter to achieve an optimal signal-to-noise ratio. However, determining the power spectral density is not feasible for all magnetic-inductive flowmeters due to the required knowledge, equipment, and time. For the user of a magnetic-inductive flowmeter, there is no easy way to select the optimal settings without conducting an extensive measurement campaign, and the manufacturer cannot specify the optimal settings either, as the noise spectrum of the application is not known in advance.Even if the power spectral density has been determined, there is no guarantee that it will remain constant over time, as a process can change. Furthermore, without knowledge of the underlying noise, it can be difficult to distinguish noise in the calculated flow velocity from actual changes in the process. This can lead to the selection of suboptimal filters. That is, either a good signal is filtered too strongly and the actual changes in the flow are attenuated, or it is filtered too little and the noise is interpreted as genuine changes in the flow velocity-dependent measured quantity.

[0005] The invention is based on the objective of solving this problem.

[0006] The problem is solved by the method according to claim 1 and the magnetic-inductive flow meter according to claim 15.

[0007] The inventive method for configuring a magnetic-inductive flow meter, wherein the magnetic-inductive flow meter comprises:

[0008] - a measuring tube for guiding a medium;

[0009] - at least two measuring electrodes and measuring electronics electrically connected to the at least two measuring electrodes for determining a flow velocity-dependent measured quantity;

[0010] - a magnetic field-generating device for generating a magnetic field that penetrates at least section by section of the measuring tube; comprises the following process steps:

[0011] - Determining a first measure of dispersion from a flow velocity-dependent first measurement signal, wherein the first measurement signal is recorded with a first integration time,

[0012] - Determining a second measure of dispersion from a second measurement signal dependent on the flow velocity, wherein the second measurement signal is recorded with a second integration time, where the first integration time differs from the second integration time;

[0013] - Determining a test value b depending on the first and second measure of dispersion values.

[0014] The invention takes advantage of the fact that the noise of the measured signal can depend on the integration time set to acquire the measured values. If the noise is not white noise, but, for example, brown noise, the noise of the two measured signals will differ. Based on the two dispersion measure values ​​characterizing the noise, a test value b characterizing the type of noise can be determined. If, on the other hand, the noise is white noise, the noise of the measured signals does not depend on the integration time. The two determined dispersion measure values ​​are then essentially identical, i.e., the second dispersion measure value lies within an acceptance range around the first dispersion measure value, or vice versa. If the test value b, and thus the type of noise, is known, settings in the or...Make configurations on the magnetic-inductive flowmeter that improve measurement performance.

[0015] Advantageous embodiments of the invention are the subject of the dependent claims.

[0016] One design envisions:

[0017] - Determining a noise value a assigned to the test value b and the first and / or second measure of dispersion.

[0018] The key to this design lies in the understanding that by determining the noise value a from the test value b and the first and / or second measure of dispersion, the underlying characteristic of the noise's spectral power density can be extracted. Once the noise value a is known, the actual noise of the process can be unambiguously characterized. This allows for a more precise estimation of the signal-to-noise ratio and thus a more accurate estimation of the uncertainty (e.g., the standard deviation) of the determined flow rates. This can help to distinguish noise from actual changes in the flow velocity-dependent measured quantity.

[0019] If the behavior of the power spectral density is known, the noise distribution can be determined. This can be done using a lookup table in which the distribution parameters are stored as a function of the noise value a, the test value b, a calibration factor, an integration period, and / or a measurement period. Alternatively, this information can be stored as functions in the electronic memory of the magnetic-inductive flowmeter. In this way, the expected unique standard deviation of the noise for the precise adjustment of the flowmeter can be determined using the noise value a and the test value b.

[0020] If the actual measured standard deviation of the signal is greater than the expected standard deviation, this could indicate that the flow rate is actually changing. This information could then be used, for example, to adjust the length of an adaptive filter.

[0021] Furthermore, the test value b and / or the noise value a, or a derived quantity, can be used to determine optimal filter settings for a Kalman filter. The expected standard deviation can also be used to calculate a signal-to-noise ratio, which in turn can be used and output as an indicator of flow reliability.

[0022] One design envisions:

[0023] - Defining a measurement period and / or an integration time depending on the noise value a and / or the test value b.

[0024] One design envisions:

[0025] - Determining a damping value depending on the noise value a and / or the test value b.

[0026] An output (e.g., the current output) of the magnetic-inductive flowmeter can depend on the attenuation value. The output signal (e.g., a current signal) can be attenuated to a greater or lesser degree depending on the noise value a and / or the test value b.

[0027] One design envisions:

[0028] - Determining the filter size of an adjustable filter depending on the noise value a and / or the test value b.

[0029] The filter size can be, for example, the length of a median filter, a binomial filter, a Gaussian filter, a mean filter, or a similar filter.

[0030] One embodiment provides that the test value b corresponds to an exponent of a frequency-dependent function (e.g. 1 / f) to describe at least a sub-area of ​​a power spectral density.

[0031] The test value b provides information regarding the application. If the medium is water (e.g., used in the cleaning cycle of the pipeline), the test value b will have a different value, one that would be obtained, for example, in pulp and paper or food and beverage applications. This can be used to monitor the cleaning process.

[0032] One embodiment provides that the noise value a corresponds to a slope of the frequency-dependent function for describing the power spectral density.

[0033] One implementation stipulates that for a quotient q of the first and second integration times, q > 2, in particular q > 5, and preferably q < 10. To achieve better distinguishability, it is advantageous for the quotient q to be as large as possible. However, especially with short measurement periods (e.g., in applications where changes need to be detected early), the maximum value for the quotient q is limited.

[0034] One design provides that the first and / or second integration time is shorter than an integration time set at the factory in the magnetic-inductive flowmeter.

[0035] This has the advantage that the configuration can take place without disturbing the flow measurement.

[0036] One embodiment provides that a distinction is made between influences on the measurement signal caused by flow velocity and those caused by background noise by means of the noise value a and / or the test value b.

[0037] One design provides that, depending on the noise value a and / or the test value b, the presence of bubbles and / or solids in the medium is determined.

[0038] One embodiment provides that the process steps, in particular continuously, take place in a measuring mode of the magnetic-inductive flowmeter.

[0039] To configure a magnetic-inductive flowmeter with an optimal signal-to-noise ratio, the underlying noise spectrum must be known. The objective of the present invention is to enable a rapid measurement with which a characteristic quantity or quantities of the power spectral density can be determined without interrupting the flow measurement. Knowing these quantities allows the optimal configuration to be selected during the operation of the magnetic-inductive flowmeter.

[0040] One embodiment provides that the magnetic-inductive flowmeter includes operating electronics for operating the magnetic field-generating device, wherein the operating electronics are configured to provide an operating signal to the magnetic field-generating device, the operating signal having an amplitude value and / or a time value which is determined as a function of the noise value a and / or the test value b, in particular continuously.

[0041] The operating signal can consist of a time-varying (e.g., alternating) voltage signal. This voltage signal can assume different voltage values ​​(amplitude values) in steps. Voltage signals are known for magnetic-inductive flowmeters in which rectangular voltage pulses with different amplitude values ​​are used, applied to the magnetic field-generating device for varying durations. The voltage signal can include a holding voltage, which is applied for a specified holding time. During the holding time, particularly at the end, the magnetic field stabilizes and is essentially constant. Therefore, the measured values ​​of the electrical potential acquired during the holding time are used to determine the flow rate. To achieve a faster stabilization of the magnetic field, a shot period can be provided before the holding time, during which a shot voltage is applied.The shot voltage is typically many times higher than the holding voltage, while the shot time is many times shorter than the holding time. The shot voltage and / or the holding voltage can be adjusted to the prevailing process conditions based on the determined noise value a and / or test value b, or a value derived from them. Alternatively, the shot time and / or the holding time can also be adjusted to the prevailing process conditions based on the determined noise value a and / or test value b, or a value derived from them.

[0042] One embodiment provides that the magnetic field-generating device comprises at least one coil, comprising:

[0043] - Setting a coil current setpoint depending on the noise value a and / or the test value b.

[0044] The magnetic-inductive flow meter according to the invention for determining a flow velocity-dependent measured quantity comprises:

[0045] - a measuring tube for guiding a medium;

[0046] - at least two measuring electrodes and measuring electronics electrically connected to the at least two measuring electrodes for determining a flow velocity-dependent measured quantity;

[0047] - a magnetic field-generating device for generating a magnetic field penetrating the measuring tube at least section by section, characterized in that the measuring electronics are configured to carry out the method according to one of the preceding claims.

[0048] The invention is explained in more detail with reference to the following figures. They show:

[0049] Fig. 1: a cross-section through a magnetic-inductive flow meter;

[0050] Fig. 2: the time response of the measurement signal;

[0051] Fig. 3: the time behavior of the flow signal determined from the measurement signal;

[0052] Fig. 4: a power spectral density; Fig. 5a, 5b: two noise ratio test result diagrams;

[0053] Fig. 6a, 6b: two noise-to-noise ratio diagrams; and

[0054] Fig. 7: a process flow of one embodiment of the method according to the invention.

[0055] Fig. 1 shows a cross-section through a magnetic-inductive flowmeter 1 known from the prior art for determining a flow velocity-dependent measured quantity. The flow velocity-dependent measured quantity is typically the current flow velocity of the flowable and electrically conductive medium to be monitored. Alternatively, the flow velocity-dependent measured quantity can be a volumetric flow rate or – if the medium density is known – a mass flow rate.

[0056] The magnetic-inductive flowmeter 1 generally comprises a magnetic field-generating device 7 for generating a magnetic field that penetrates at least a section of the measuring tube. The generated magnetic field itself has an imaginary principal axis HA that lies within an imaginary plane of symmetry of the generated magnetic field distribution and / or an imaginary plane of symmetry of the measuring tube 8. The magnetic field-generating device 7 can comprise at least one magnetic coil. Magnetic-inductive flowmeters with two opposing magnetic coils are known. Alternatively or additionally, the magnetic field-generating device 7 can (additionally) include a permanent magnet for generating a magnetic field. The at least one magnetic coil itself can be a cylindrical coil or a saddle coil. Furthermore, magnetic field-guiding components can be part of the magnetic field-generating device 7.The magnetic field-generating device 7 can thus have at least one coil core, at least one pole shoe, and / or at least one field-guiding component (e.g., a field-guiding plate) for guiding the magnetic field outside the medium. The magnetic field-generating device 7 is electrically connected to operating electronics 10, which are configured to provide an operating signal (e.g., a voltage signal with time-varying voltage amplitudes) to the magnetic field-generating device 7.

[0057] The magnetic-inductive flowmeter 1 further comprises a measuring tube 8 for guiding the medium to be monitored. The measuring tube 8 typically consists of a metallic support tube whose inner surface is provided with an electrically insulating liner 2. The liner 2 is in contact with the medium. Alternatively, the measuring tube 8 can also have a support tube made directly of an electrically insulating material (e.g., plastic or ceramic), thus eliminating the need for a liner 2. The measuring tube 8 can have a round cross-section, as shown, or—at least in the measuring section—a rectangular cross-section. To clearly locate the individual electrodes on the measuring tube 8, the measuring tube 8 is subsequently divided into imaginary subsections. An imaginary horizontal first longitudinal plane LE1 divides the measuring tube 1 into a first and second subsection TA1 and TA2.The first section TA1 is the upper half of measuring tube 1, and the second section TA2 is the lower half of measuring tube 1. The two sections TA1 and TA2 are essentially the same size. The first longitudinal plane LE1 is intersected perpendicularly by the principal axis HA.

[0058] In principle, a magnetic-inductive flowmeter 1 according to the prior art has at least two opposing measuring electrodes 3, 4. However, magnetic-inductive flowmeters with more than one measuring electrode on each side of the measuring tube are also known. The at least two measuring electrodes 3, 4 are intersected by the imaginary longitudinal plane LE1. The illustrated prior art further shows a level monitoring electrode 5, which is arranged in the first subsection TA1, and a reference electrode 6, which is arranged in the second subsection. The reference electrode 6 is connected to an electrical reference potential. The function of a level monitoring electrode 5 includes the detection of the fill level in a measuring tube 8.Previously, it was possible, among other things, to measure the electrical resistance between the level monitoring electrode 5 and a reference electrode 6 or the process connection, which may be equipped with grounding discs or ground electrodes. The measured electrical resistance or conductivity increases abruptly when the measuring tube 8 changes from a fully filled to a partially filled state. In this case, air would be present between the level monitoring electrode 5 and the reference electrode 6, at least in some areas, so that the conductivity of air would be included in the measured electrical resistance or conductivity. The corresponding change at the level monitoring electrode 5 is detected by the measuring electronics and displayed by an output unit (e.g., a display).A display unit, a signal lamp, or a connector outputs a signal regarding the fill level of the measuring tube 8. The level monitoring electrode 5 is also referred to in the literature as an empty-pipe detection electrode, or EPD electrode for short. The function of a reference electrode 6 is to ensure potential equalization between the medium and the sensor (measuring tube, electrodes, etc.). The reference electrode 6 provides an alternative grounding option to the conventional grounding disc. Furthermore, the potential difference between measuring electrodes 3, 4, and reference electrode 6 can be used for evaluating the flow rate, for example, for analyzing the flow profile. The reference electrode 6 is typically electrically connected to the housing (not shown) of the transmitter unit (not shown) and / or the pipeline (not shown). The housing is typically connected to the protective earth.Magnetic-inductive flowmeters are already available on the market whose reference electrode 6 is not connected to ground. This means that the reference electrode is not directly connected to an external reference potential. All four electrodes 3-6 shown are electrically connected to a measuring electronics unit 9. The measuring electronics unit 9 is configured to measure a measuring voltage across the two measuring electrodes 3 and 4 and to determine a flow velocity-dependent measured quantity from this. Furthermore, the measuring electronics unit 9 is configured to perform a conductivity measurement between the level monitoring electrode 5 and one of the other electrodes (e.g., the reference electrode 6) in order to detect partial filling. The measuring electronics unit 9 is also electrically connected to the magnetic field-generating device 7 and configured to provide an operating signal to the magnetic field-generating device 7.The operating signal can be a time-varying voltage signal. The measuring electronics 9 can comprise at least a printed circuit board, a microprocessor, electronic components (e.g., electrical resistors, capacitors, logic electronic components, power supply, filters, etc.), conductive traces, and / or connectors. The same applies to the operating electronics 10.

[0059] Commercially available magnetic-inductive flowmeters 1 typically have a transmitter unit (not shown) with a display. The transmitter unit can be directly connected to the housing (not shown) that encloses the measuring tube 8. In this case, with the magnetic-inductive flowmeter in its usual mounting orientation, the transmitter unit would be located in the upper part of the flowmeter.

[0060] Alternatively, the transmitter unit can be arranged separately from the measuring tube 8 and housing, allowing the operator to position it almost anywhere from the pipeline. The measuring electronics 9 or the magnetic field-generating device 7 and / or the electrodes 3-5 are then connected via cables to a sensor output (e.g., a connector) in the housing. The sensor output itself is then also connected, or can be connected, to the remote transmitter unit via a cable.

[0061] Fig. 2 shows the time response of the individual measured values ​​of the measurement signal within 300 milliseconds. The measuring electronics are designed to record or determine the flow velocity-dependent electrical potential (here in mV) applied to the measuring electrodes or the potential difference, for example, between two measuring electrodes, and to calculate a flow velocity-dependent measured quantity from this – in this case, the flow velocity itself (see Fig. 3). The switching of the magnetic field is reflected in the measurement signal. The measurement signal alternates at a frequency of approximately 25 Hz by an offset of approximately 55.5 mV. Each individual measured value of the measurement signal is itself an average value, determined from a large number of individual measured values, which are integrated or summed within an integration time. The integration time is typically a few milliseconds.

[0062] Figure 3 shows the time response of the flow signal derived from the measurement signal. The flow signal shown here is the flow velocity in m / s. The flow signal was recorded over a period of 30 seconds. The flow signal itself comprises a multitude of flow measurements. The measured value of the electrical potential (or the potential difference) used to determine the flow measurements is calculated by averaging a large number of individual electrical potential measurements over the integration time. In fact, the individual electrical potential measurements are taken at a specific sampling rate. Furthermore, a measure of dispersion (e.g., the standard deviation) of the measured values ​​from the mean can be used as a measure of noise.

[0063] Figure 4 shows a power spectral density of the measurement signal as a function of frequency. Frequency and power density are plotted logarithmically. The power spectral density of a measurement signal can be determined, for example, by performing a Fourier transform on the signal. The power spectral density shown was determined using a computer algorithm from the unfiltered measurement signal of a so-called "pulp-and-paper" application. The measurement period of 20 ms (two 20 ms measurement periods correspond to a frequency of 25 Hz) is clearly identifiable from the power spectral density. The mains frequency of the power supply can also be identified in the power spectral density.

[0064] The spectral power density of a measurement signal is defined as the power distributed across a specific frequency range, divided by that range. The spectral power density is a mathematical function of frequency. The integral of the spectral power density over all frequencies yields the total power of the measurement signal.

[0065] The type of noise can be determined from the power spectral density. It is assumed that the spectral power density follows a simple power law: PSD = a - f b The following follows, where f is the frequency, a is the noise value, and b is the test value. To determine the noise value a and the test value b, the power function is fitted to the power spectral density. This shows that the power spectral density is at least piecewise linearly proportional to 1 / with a slope of -2. Thus, brown noise (test value b = 2) is present, with a noise value a of 2.8 × 10⁻⁸ V. 2 / Hz.

[0066] As explained above, such a procedure is extremely cumbersome and not simply possible for existing magnetic-inductive flow meters.

[0067] Figures 5a and 5b show two noise ratio-test value diagrams. The diagram in Figure 5a shows the noise ratio of the noise (e.g., standard deviation from mean) of a measurement signal recorded with an integration time of 2 milliseconds and the noise of a measurement signal recorded with an integration time of 1 millisecond. The diagram in Figure 5b shows the noise ratio of the noise of a measurement signal recorded with an integration time of 5 milliseconds and the noise of a measurement signal recorded with an integration time of 1 millisecond. The noise ratios were calculated for different test values ​​b and different noise values ​​a (1 × 10⁻⁷ to 1 × 10⁻¹⁰ V). 2 / Hz). Alternatively, the values ​​can also be determined experimentally. The simulations yield two functions (Figs. 5a and 5b) that represent the correlation between the noise ratio and the test value b. It is evident that there are no significant differences between the simulated noise ratios for different noise values ​​a. The core idea of ​​the invention is that the relationships between the noise ratios of different integration times and the test values ​​b are used to optimally configure the magnetic-inductive flowmeter. The simulated functions or simulation values ​​can be stored in an electronic memory within the magnetic-inductive flowmeter, particularly in the measuring electronics. Taking these into account in conjunction with the determined current noise ratios, the test value b can be calculated.

[0068] The specific procedure is illustrated below using two examples. In each case, a measurement signal is acquired with a first integration time (2 ms or 5 ms) and a second measurement signal with a second integration time (1 ms). From the two measurement signals, the noise (e.g., the standard deviation from the mean), i.e., two measures of dispersion, are determined. The noise ratio of a first measure of dispersion, acquired with an integration time of 2 ms, to a second measure of dispersion, acquired with an integration time of 1 ms, is approximately 1.49. In the second example, the noise ratio of a first measure of dispersion, acquired with an integration time of 5 ms, to a second measure of dispersion, acquired with an integration time of 1 ms, is approximately 2.36. As can be seen in Figures 5a and 5b, this corresponds to a test value b of 2.The test value b is derived from the x-coordinate of the intersection point of the dotted line (the y-value corresponds to the determined ratio) and the respective simulated function. Or from the intersection point of the dotted line with the solid line.

[0069] Figures 6a and 6b each show five linear functions that define the noise or a measure of dispersion for a specific integration time (here 1 ms and 5 ms) as a function of the noise value a. Each linear function was simulated for a given test value b. The measure of dispersion increases with decreasing test value b. The lowest linear function results from a test value b of 2, and the highest linear function results from a test value b of 0. Using the known linear relationship, the known measure of dispersion, and taking into account the previously determined test value b (in this case, b = 2), the noise value a can be determined. From the measured noise (approx. 8E-6 V), which is obtained with an integration time of 1 ms, a noise value a of 3.3E-8 is obtained, and from the measured noise (approx. 2E-5 V), which is obtained with an integration time of 5 ms, a noise value a of 3.2E-8 is obtained.Both values ​​are close to the measured noise value a of 2.8 × 10⁻⁸, see Fig. 4. Once the test value b and / or the noise value a are known, several possibilities arise. One possibility would be to use predefined optimal settings for the measurement period and / or the integration time as a function of the noise value a or the test value b to enable automatic adjustment of the optimal configuration. It would also be possible to change this dynamically if the underlying noise were to change.

[0070] Knowing the noise value α allows for an accurate estimation of the signal-to-noise ratio, thus enabling estimates of the uncertainty in the measured flow rates. This can help distinguish noise from actual changes in flow rate.

[0071] The previously described approach allows the characteristics of the underlying spectral power density to be determined. One advantage is that the noise can be easily evaluated for multiple integration times, and as long as these are all shorter than the configured integration time, this has no impact on the operation of the magnetic-inductive flowmeter.

[0072] Fig. 7 shows a process flow of an embodiment of the inventive method for configuring a magnetic-inductive flow meter (see Fig. 1).

[0073] In process step I, a first measure of dispersion is determined from a flow velocity-dependent initial measurement signal. This first measure of dispersion can be a standard deviation, a variance, or a mean deviation from the mean of a measurement signal. The initial measurement signal can be a raw measured signal or a signal derived from the raw signal. The initial measurement signal can comprise a set of measured values ​​acquired sequentially over time. These measured values ​​could, for example, be voltage values.

[0074] According to the invention, the first measurement signal is acquired with a first integration time. The integration time is defined as the time in which measured raw values ​​(individual values) are acquired before they are calculated (e.g., to obtain a mean or other statistical value).

[0075] In process step II, a second measure of dispersion is determined from a second measurement signal that is dependent on the flow velocity. As with the first measure of dispersion, the second measure of dispersion can be a standard deviation, a variance, or a mean deviation from the mean value of a measurement signal. The second measurement signal can be a raw measured signal or a signal derived from the raw signal. The second measurement signal can comprise a set of measured values ​​that were recorded sequentially. These measured values ​​could, for example, be voltage values.

[0076] Unlike the first measurement signal, the second measurement signal was acquired with a second integration time. The first integration time differs from the second integration time. The integration time was chosen so that it does not influence the measurement.

[0077] The first and second integration times can be chosen, for example, such that for a quotient q of the first and second integration times, q > 2, in particular q > 5 and preferably q < 10.

[0078] The first integration time can be factory-set, while the second integration time is shorter than the first. Measurements using a second integration time can therefore be performed during noise analysis phases. In a normal measurement phase, however, measurements are taken using the first integration time.

[0079] In process step III, a test value b is determined as a function of the first and second measures of dispersion. The test value b is a noise-specific quantity; that is, it indicates the type of noise present. In a power spectral density, the test value b is derived from the exponent of a frequency-dependent function (e.g., 1 / f) describing the power spectral density, or, in a log-log plot, from the slope of the linear portion of the power density. However, according to the invention, the test value b is not determined from the power spectral density, but rather from a stored relationship between the test value b and the first and second measures of dispersion. This relationship can be determined experimentally or simulated. It can also be described by a mathematical function, which is itself stored in the measurement electronics.Alternatively, only individual values ​​of the relationship can be stored, which can then serve as the basis for an interpolation procedure if needed.

[0080] In process step IV, a noise value a is determined, which is assigned to the test value b and the first and / or second measure of dispersion. The noise value a is the offset of the linear function describing the power spectral density in the log-log representation. Alternatively, the noise value a can also correspond to the slope of the frequency-dependent function (1 / f) describing the power spectral density. However, according to the invention, the noise value a is not determined from the power spectral density, but rather as a function of the first and / or second measure of dispersion and taking into account the previously determined test value b. The relationship between the test value b, the noise value a, and the first and / or second measure of dispersion can be determined experimentally or simulated. The relationship can be described by a mathematical function (e.g., a straight line), which in turn is itself stored in the measuring electronics.Alternatively, only individual values ​​of the relationship can be stored, which can then serve as the basis for an interpolation procedure if needed.

[0081] In a process step V, a measurement period and / or an integration time is determined depending on the noise value a and / or the test value b.

[0082] In high-noise applications, magnetic inductive flowmeters with a stronger signal (i.e., a stronger magnetic field) and higher measurement frequencies are often used. When configuring a magnetic inductive flowmeter, it is usually possible to set a measurement period, which defines the time between measurements, an integration time, during which the recorded signal is integrated, and sometimes even the coil current. The minimum measurement time is influenced by the time required to induce the magnetic field, and a stronger signal results in a longer measurement time. Furthermore, choosing a longer integration time limits the minimum possible measurement duration. The selection of measurement time and integration time settings that yields the best signal-to-noise ratio depends on the spectral power density of the noise in a given application, and this is generally unknown.Instead, the magnetic-inductive flowmeter is usually configured based on experience and time-consuming trials, or often simply left at the factory settings.

[0083] Alternatively or additionally, the following procedural steps may be required.

[0084] This allows an adjustable damping value to be set depending on the noise value a and / or the test value b.

[0085] Additionally or alternatively, the filter size of an adjustable filter in the measuring electronics can be set depending on the noise value a and / or the test value b. The adjustable filter could be, for example, a Kalman filter or median filter.

[0086] Additionally or alternatively, the noise value a and / or the test value b can be used to differentiate between flow velocity-related and background noise-related influences on the measurement signal. This information can be displayed on a display unit. For example, the operator can be informed that the process is causing increased noise.

[0087] Additionally or alternatively, the presence of bubbles and / or solids in the medium can be determined depending on the noise value a and / or the test value b. This can be used to identify a process change (e.g., cleaning with water). The above-mentioned process steps can be carried out, particularly continuously, in a single measurement mode of the magnetic-inductive flowmeter.

[0088] Additionally or alternatively, depending on the noise value a and / or the test value b, an amplitude value (e.g. a holding voltage and / or a shot voltage) and / or a time value (e.g. a holding time and / or a shot time) of the operating signal, in particular continuously, can be determined.

[0089] Additionally or alternatively, a coil current setpoint can be defined depending on the noise value a and / or the test value b. Control methods are known in which the operating signal is adjusted to achieve a coil current setpoint. This requirement can be linked to an additional condition. For example, it may be required that the coil current setpoint be reached within a specific time interval. The core concept of this design is that the coil current setpoint is not chosen to be constant, but is adjustable depending on the noise value a and / or the test value b, or a quantity derived from them.

[0090] REFERENCE MARK LIST

[0091] 1 magnetic-inductive flow meter

[0092] 2 liner 3 measuring electrode

[0093] 4 measuring electrode

[0094] 5 Level monitoring electrode

[0095] 6 Reference electrode

[0096] 7 magnetic field generating device 8 measuring tube

[0097] 9 Measuring electronics

[0098] 10 Operating electronics

Claims

PATENT CLAIMS 1. Method for configuring a magnetic-inductive flowmeter (1), wherein the magnetic-inductive flowmeter (1) comprises: - a measuring tube (8) for guiding a medium; - at least two measuring electrodes (3, 4) and measuring electronics (9) electrically connected to the at least two measuring electrodes (3, 4) for determining a flow velocity-dependent measured quantity; - a magnetic field-generating device (7) for generating a magnetic field that penetrates the measuring tube (8) at least partially; comprising the process steps: - Determining a first measure of dispersion from a flow velocity-dependent first measurement signal, wherein the first measurement signal is recorded with a first integration time, - Determining a second measure of dispersion from a second measurement signal dependent on the flow velocity, wherein the second measurement signal is recorded with a second integration time, where the first integration time differs from the second integration time; - Determining a test value b depending on the first and second measure of dispersion values.

2. The method of claim 1, comprising: - Determining a noise value assigned to the test value and the first and / or second measure of dispersion value a.

3. The method of claim 1 or 2, comprising: - Defining a measurement period and / or an integration time depending on the noise value a and / or the test value b.

4. A method according to any one of the preceding claims, comprising: - Determining a damping value depending on the noise value a and / or the test value b.

5. A method according to any one of the preceding claims, comprising: - Determining the filter size of an adjustable filter depending on the noise value a and / or the test value b.

6. Method according to one of the preceding claims, wherein the test value b corresponds to an exponent of a frequency-dependent function for describing at least a sub-area of ​​a power spectral density.

7. Method according to any one of claims 2 to 6, where the noise value a corresponds to a slope of the frequency-dependent function for describing the power spectral density.

8. Method according to one of the preceding claims, wherein for a quotient q of the first and second integration time, q > 2, in particular q > 5 and preferably q < 10.

9. Method according to any of the preceding claims, wherein the first and / or second integration time is less than an integration time set at the factory in the magnetic-inductive flowmeter (1).

10. Method according to one of the preceding claims, wherein a distinction is made between flow velocity-related and background noise-related influences on the measurement signal by means of the noise value a and / or the test value b.

11. Method according to one of the preceding claims, wherein the presence of bubbles and / or solids in the medium is determined depending on the noise value a and / or the test value b.

12. Method according to one of the preceding claims, wherein the method steps, in particular continuously, are carried out in a measuring operation of the magnetic-inductive flowmeter (1).

13. Method according to one of the preceding claims, wherein the magnetic-inductive flowmeter (1) comprises operating electronics (10) for operating the magnetic field-generating device (7), wherein the operating electronics (10) is configured to provide an operating signal to the magnetic field-generating device (7), wherein the operating signal has an amplitude value and / or a time value which is determined as a function of the noise value a and / or the test value b, in particular continuously.

14. Method according to any of the preceding claims, wherein the magnetic field-generating device (7) comprises at least one magnetic coil, comprising: - Setting a coil current setpoint depending on the noise value a and / or the test value b.

15. Magnetic-inductive flow meter for determining a flow velocity-dependent measured quantity, comprising: - a measuring tube (8) for guiding a medium; - at least two measuring electrodes (3, 4) and measuring electronics (9) electrically connected to the at least two measuring electrodes (3, 4) for determining a flow velocity-dependent measured quantity; - a magnetic field-generating device (7) for generating a magnetic field penetrating the measuring tube (8) at least section by section, characterized in that the measuring electronics (9) is configured to carry out the method according to one of the preceding claims.

Images