JP5974518B2 - Coriolis mass flow meter - Google Patents

Description

  The present invention relates to a Coriolis mass flow meter for measuring a flow rate of a fluid flowing through a pipeline.

  The Coriolis mass flow meter measures the flow rate of fluid flowing through a pipeline. For this purpose, the pipeline is vibrated by vibration. The vibration signal is detected by the upstream coil provided on the upstream side of the pipe and the downstream coil provided on the downstream side, and the flow rate of the fluid flowing through the pipe is calculated based on the phase difference between the two vibration signals. To measure the flow rate.

  This type of Coriolis mass flow meter is disclosed in Patent Document 1. Particularly in FIG. 5 of Patent Document 1, an upstream coil signal (vibration signal detected by the upstream coil), a downstream coil signal (vibration signal detected by the downstream coil), and a reference signal are branched into a plurality of paths, The calculation of the phase difference between the upstream coil signal and the downstream coil signal and the calculation of the phase difference between the reference signals are performed in parallel by switching the path of each signal.

  In each path after branching, a phase shift occurs due to a change in the environmental condition of the signal path. In order to compensate for this phase shift, the zero compensation value is calculated by calculating the phase difference between the reference signals. The measurement error is reduced by correcting the phase difference between the upstream coil signal and the downstream coil signal with a zero compensation value.

  And since the calculation of the zero compensation value and the calculation of the phase difference for the flow measurement are performed in a time-sharing manner, the time during which the flow measurement is interrupted can be shortened, and the stable flow measurement It can be performed. In addition, since the zero compensation value is calculated using the reference signal, stable flow rate measurement can be performed without causing frequency fluctuation.

JP 2011-137771 A

  Since the technique of Patent Document 1 branches the reference signal and calculates the zero compensation value using the phase difference, it is possible to reduce an error due to the phase shift of the path. In addition, the phase difference calculation for the zero compensation value and the phase difference calculation for the flow rate measurement are performed in parallel, and the zero compensation value is calculated using a reference signal that does not cause frequency fluctuations. This is a technique that has an extremely excellent effect in that the flow rate can be measured.

  By the way, in the technique of Patent Document 1, the upstream coil signal, the downstream coil signal, and the reference signal are branched and switched. Therefore, a plurality of signal paths are arranged in parallel. For example, in FIG. 5 of Patent Document 1, four signal paths (channels) are input to the input amplifier by the switching circuit.

  At this time, the signals of the respective channels may be superimposed on each other. An error occurs in the flow rate measurement by superimposing the signal of the other channel on the signal of each channel. Therefore, the accuracy of flow rate measurement is reduced.

  In the technique of Patent Document 1, the reference signal generation circuit generates a reference signal, but the reference signal uses a signal of an arbitrary frequency. Therefore, the frequency (natural frequency) of the upstream coil signal and the downstream coil signal may coincide with the frequency of the reference signal (reference signal frequency). When the natural frequency and the reference signal frequency are the same frequency, the superimposed component of the signal becomes a constant component (offset component), and it is extremely difficult to remove and separate the superimposed component from each signal. As a result, an error occurs in the calculation of the flow rate measurement and the calculation of the zero compensation value performed using the signal of each channel.

  Therefore, the present invention has an object to reduce measurement error even when signal superimposition occurs when a flow signal is measured by branching a vibration signal and a reference signal into a plurality of paths and calculating a phase difference. .

In order to solve the above problems, the Coriolis mass flowmeter of the present invention vibrates a pipeline through which a fluid flows, and based on the first vibration signal on the upstream side and the second vibration signal on the downstream side of the pipeline, A Coriolis mass flow meter for measuring a flow rate of a fluid flowing through the pipe, and generating a reference signal having a frequency different from the frequencies of the first vibration signal and the second vibration signal, which are natural frequencies of the pipe A reference signal generator, and a path of the first vibration signal, the second vibration signal, and the reference signal is branched into a plurality of phases, and a phase difference between the first vibration signal and the second vibration signal is branched. a circuit error calculation unit that calculates a circuit error and the phase difference calculation unit, in the natural frequency by based rather approximate calculation of the phase difference between the reference signal for calculating a phase difference between the reference signal, the phase difference Calculated by the calculation unit A circuit error correction unit with respect to the phase difference between the first vibration signal and the second oscillation signal corrects the circuit error, the circuit error the second oscillating signal and the corrected first oscillation signal by the correction unit and A flow rate calculation unit for calculating the flow rate based on the phase difference of
It is provided with.

  According to this Coriolis mass flow meter, the frequency of the reference signal is set to a frequency different from that of the first vibration signal and the second vibration signal. As a result, even if signals are superimposed, they are not superimposed as constant components. As a result, it is possible to remove the superimposed component and reduce the measurement error.

  In addition, a filter is provided for removing a superimposed component when each signal transmitted through the branched path is superimposed on a signal transmitted through another path.

  Since the frequency of the reference signal is different from the frequencies of the first vibration signal and the second vibration signal, it is possible to remove the superimposed component even if a superimposed component occurs. Therefore, a measurement error can be further reduced by providing a filter that removes the superimposed component.

  The reference signal generation unit outputs two or more reference signals having different frequencies in a time-sharing manner, and the circuit error calculation unit calculates the circuit error by a first-order approximation or more. And

  Although reference signals having different frequencies are used to calculate the circuit error, the relationship between the circuit error and the frequency is not linearly proportional. Therefore, by calculating the circuit error by approximation of the first order or more using two or more reference signals, it is possible to improve the calculation accuracy of the circuit error.

  In addition, the frequency of the reference signal is determined based on an approximation error when the circuit error is calculated by the circuit error calculation unit and an effect of removing the superimposed component by the filter.

  When the difference between the frequency of the first vibration signal and the second vibration signal and the frequency of the reference signal increases, the calculation accuracy of the circuit error by approximation decreases. On the other hand, if the frequency difference becomes too small, it is difficult to remove the superimposed component. Therefore, by determining the frequency of the reference signal based on the approximate error when calculating the circuit error and the effect of removing the superimposed component, the accuracy of calculating the circuit error is improved and the removal of the superimposed component is improved. It becomes possible.

  In addition, a reference signal frequency storage unit that stores a frequency different from the frequencies of the first vibration signal and the second vibration signal, and a frequency stored in the reference frequency storage unit are selected, and the reference signal generation unit And a reference signal frequency selection unit that outputs the signal.

  By storing in advance a frequency different from the frequencies of the first vibration signal and the second vibration signal, a reference signal having a fixed frequency can be output. For this reason, the frequency of the reference signal can be selected without requiring complicated control.

  Further, the calculation of the phase difference between the first vibration signal and the second vibration signal and the calculation of the phase difference between the reference signals are performed in parallel.

  By performing the phase difference calculation for measuring the flow rate and the phase difference calculation for obtaining a circuit error in parallel, the number of branches of each signal increases. As a result, the number of signal paths (channels) increases and signal superposition is likely to occur, but the superposition component can be removed, so that errors in flow rate measurement can be reduced. In addition, since the calculation is performed in parallel, the time for interrupting the flow rate measurement can be shortened.

  In the present invention, the frequency of the reference signal is set to a frequency different from that of the first vibration signal and the second vibration signal. Thus, even if signals are superimposed, they are not superimposed as a constant component, so that the superimposed component can be removed. For this reason, a measurement error can be reduced.

It is a block diagram which shows the structure of a Coriolis mass flowmeter. It is a block diagram which shows the structure of a digital signal processing circuit. It is a block diagram which shows the structure of a phase difference calculation part. It is a graph which shows the relationship between a frequency and a circuit error. It is a block diagram which shows the structure of the other example of a digital signal processing circuit. It is a block diagram which shows the structure of the further another example of a digital signal processing circuit.

  Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 shows a Coriolis mass flow meter 1 of the present embodiment. In this embodiment, the Coriolis mass flow meter uses a digital signal processing type straight tube type Coriolis mass flow meter, but a U-shaped one may be used. The Coriolis mass flow meter 1 vibrates the pipe through which the fluid flows, and measures the flow rate of the pipe based on the phase difference between the vibration on the upstream side and the downstream side of the pipe. For example, a tube can be used as the conduit.

  As shown in FIG. 1, the Coriolis mass flow meter 1 includes a detector 2 and a converter 3. The detector 2 includes an upstream coil L1, a downstream coil L2, a drive coil L3, and a resistance temperature detector RTD (Resistance Temperature Detector). Based on the drive signal DRV output from the converter 3, the drive coil L3 excites a conduit (not shown). Thereby, the pipe line vibrates.

  The upstream coil L1 detects the vibration on the upstream side of the pipe and outputs this to the converter 3 as the first vibration signal S1. The first vibration signal S1 is an upstream coil signal. The downstream coil L2 detects the vibration on the downstream side of the pipe and outputs this to the converter 3 as the second vibration signal S2. The second vibration signal S2 is a downstream coil signal. The resistance temperature detector RTD is a temperature sensor that detects the temperature of the fluid flowing through the pipe line, and operates based on the resistance drive signal RTDDRV output from the converter 3. Then, the detected temperature is output to the converter 3 as a temperature signal TEMP.

  The converter 3 is connected to the detector 2 and measures the mass flow rate of the fluid flowing through the pipeline based on the first vibration signal S1 and the second vibration signal S2. The temperature signal TEMP is used to compensate for temperature fluctuations. The converter 3 includes a reference signal generation circuit 10, a first switching circuit 11 to a fourth switching circuit 14, a first input amplifier 15 to a fourth input amplifier 18, a phase difference ADC 19, a digital signal processing circuit 20, a CPU 21, and a drive circuit 22. An RTD driving circuit 23, a temperature ADC 24, a first output circuit 25, a second output circuit 26, and a display 27 are provided.

  The reference signal generation circuit 10 is a reference signal generation unit that generates a reference signal. This reference signal is a signal for correcting a circuit error in the path. Here, path 1 and path 2 shown in FIG. 1 are correction targets for circuit errors. The reference signal generation circuit 10 is connected to the digital signal processing circuit 20, and the frequency of the reference signal is set as the reference signal frequency fa by the digital signal processing circuit 20. Then, a reference signal having the reference signal frequency fa is output.

  The reference signal output from the reference signal generation circuit 10 is branched into four to become reference signals S301 to S304. Further, the first vibration signal S1 branches into two in the middle to become the first vibration signals S1 and S102. Further, the second vibration signal S2 branches into two on the way to become second vibration signals S2 and S201.

  The first switching circuit 11 inputs the first vibration signal S1 and the reference signal S301, and selectively switches and outputs one of the signals. The second switching circuit 12 receives the reference signal S302 and the second vibration signal S201, and selectively switches and outputs one of the signals. The third switching circuit 13 receives the first vibration signal S102 and the reference signal S303, and selectively switches and outputs one of the signals. The fourth switching circuit 14 receives the reference signal S304 and the second vibration signal S2, and selectively switches and outputs one of the signals.

  The first input amplifier 15 amplifies and outputs the signal selected by the first switching circuit 11. The second input amplifier 16 amplifies and outputs the signal selected by the second switching circuit 12. The third input amplifier 17 amplifies and outputs the signal selected by the third switching circuit 13. The fourth input amplifier 18 amplifies and outputs the signal selected by the fourth switching circuit 14.

  The path of the signal amplified by the first input amplifier 15 to the fourth input amplifier 18 is a channel. Here, there are four channels (CH1 to CH4). Signals are input to the phase difference ADC 19 from these four channels CH1 to CH4. The phase difference ADC 19 is an analog / digital converter, and converts an input analog signal into a digital signal.

  The signals of the channels CH1 to CH4 converted into digital signals are input to the digital signal processing circuit 20. The digital signal processing circuit 20 calculates the phase difference between the first vibration signal S1 (or S102) and the second vibration signal S2 (or S201). This calculated phase difference becomes a flow rate signal indicating the flow rate flowing through the pipeline. Further, the digital signal processing circuit 20 calculates a phase difference between the reference signals in order to correct a circuit error. Details of the digital signal processing circuit 20 will be described later.

  The CPU 21 calculates the flow rate (flow rate calculation) flowing through the pipe line based on the flow rate signal output from the digital signal processing circuit 20. Since the fluid flow rate causes a measurement error due to fluid temperature fluctuations, the flow rate signal is corrected based on the temperature signal TEMP measured by the RTD RTD to compensate for the measurement error due to fluid temperature fluctuations. To do. Thereby, the CPU 21 performs flow rate measurement.

  The drive circuit 22 outputs a drive signal DRV. The drive signal DRV is input to the drive coil L3, and the vibration of the pipeline is performed. The drive circuit 22 positively feeds back the first vibration signal S1 or the second vibration signal S2, and at the frequency of the first vibration signal S1 and the second vibration signal S2, that is, the natural vibration frequency (natural frequency) f0 of the pipe line. The drive coil L3 is vibrated.

  The RTD drive circuit 23 outputs a resistance drive signal RTDDRV for setting an operating current to the resistance temperature detector RTD. Thereby, the resistance temperature detector RTD measures the temperature of the fluid. The temperature measured by the resistance temperature detector RTD is input to the temperature ADC 24 as the temperature signal TEMP.

  The temperature ADC 24 is an analog-digital converter that converts the analog temperature signal TEMP into a digital signal. The temperature signal TEMP converted into a digital signal by the temperature ADC 24 is input to the CPU 21 and used to compensate for a measurement error due to fluid temperature fluctuation as described above.

  The CPU 21 measures the flow rate and outputs flow rate data indicating the flow rate of the fluid to the first output circuit 25 and the second output circuit 26. The first output circuit 25 and the second output circuit 26 output flow rate data to the outside. The flow rate data is also input to the display 27 to display the flow rate data. The display 27 can also display information such as temperature.

  In addition, as shown in FIG. 1, the areas surrounded by the two-point difference line are path 1 and path 2. The path 1 is a path that includes the first switching circuit 11, the second switching circuit 12, the first input amplifier 15, and the second input amplifier 16 and reaches the digital signal processing circuit 20. The path 2 is a path that includes the third switching circuit 13, the fourth switching circuit 14, the third input amplifier 17, and the fourth input amplifier 18 and reaches the digital signal processing circuit 20.

  FIG. 2 shows the configuration of the digital signal processing circuit 20. The digital signal processing circuit 20 includes a phase difference calculation unit 31, a first filter 32, a second filter 33, a circuit error calculation unit 34, a circuit error correction unit 35, a frequency calculation unit 36, and a reference signal frequency selection unit 37. It is composed.

  As shown in FIG. 1, the first switching circuit 11 to the fourth switching circuit 14 selectively switch one signal out of two signals. It is assumed that “′” is added to the selected signal. Therefore, the first vibration signal S1 ′ or the reference signal S301 ′ is input from the channel CH1 of the first switching circuit 11.

  Similarly, the second vibration signal S201 ′ or the reference signal S302 ′ is input from the channel CH2 of the second switching circuit 12. The first vibration signal S102 ′ or the reference signal S303 ′ is input from the channel CH3 of the third switching circuit 13. The second vibration signal S2 ′ or the reference signal S304 ′ is input from the channel CH4 of the fourth switching circuit 14.

  The phase difference calculation unit 31 inputs signals from CH1 to CH4 and calculates the phase difference of the input signals. As shown in FIG. 3, the phase difference calculation unit 31 includes a Hilbert transformer 40 and a trigonometric function calculator 43, and the Hilbert transformer 40 includes a first FIR filter 41 and a second FIR filter 42. Configured.

  The first FIR filter 41 and the second FIR filter 42 are FIR (Finite Impulse Response) filters. The signals of CH1 to CH4 are branched to the first FIR filter 41 and the second FIR filter 42 and inputted to both. The first FIR filter 41 is an in-phase digital filter that converts an output signal having the same phase as the input signal, and the second FIR filter 42 is an out-of-phase digital filter that converts an output signal having a phase that is 90 degrees different from the input signal.

  The first FIR filter 41 outputs a signal in the form of “A × sin (ωt)” to the trigonometric function calculator 43 with respect to the input signal. The second FIR filter 42 outputs a signal in the form of “A × cos (ωt)” to the trigonometric function calculator 43 with respect to the input signal.

  The trigonometric function calculator 43 calculates the phase difference. At this time, the trigonometric function calculator 43 samples the phase differences at a plurality of times and calculates the average value of the phase differences. Accordingly, the calculated phase difference is averaged. As described above, when a superimposed component is generated in signals CH1 to CH4, since the natural frequency f0 is different from a reference signal frequency fa described later, the superimposed component appears as a vibration component. Therefore, by calculating the average value of the phase differences with the trigonometric function calculator 43, the superimposed component can be removed. Therefore, the trigonometric function calculator 43 has a function as a filter that removes the superimposed component.

  The trigonometric function calculator 43 calculates the phase difference between the first vibration signal S1 ′ and the second vibration signal S201 ′. The first vibration signal S1 ′ and the second vibration signal S201 ′ are signals transmitted through the path 1, and the phase difference is calculated. This is shown in FIG. 3 as “S1′−S201 ′ phase difference” (path 1).

  Similarly, the trigonometric function calculator 43 calculates the phase difference between the first vibration signal S102 ′ and the second vibration signal S2 ′. The first vibration signal S102 ′ and the second vibration signal S2 ′ are signals transmitted through the path 2, and their phase differences are calculated. This is shown as “S102′−S2 ′ phase difference” (path 2) in FIG.

  The trigonometric function calculator 43 calculates the phase difference between the reference signal S301 ′ and the reference signal S302 ′. The reference signal S301 ′ and the reference signal S302 ′ are signals transmitted through the path 1, and their phase differences are calculated. This is shown as “S301′-S302 ′ phase difference” (path 1) in FIG.

  The trigonometric function calculator 43 calculates a phase difference between the reference signal S303 ′ and the reference signal S304 ′. The reference signal S303 ′ and the reference signal S304 ′ are signals transmitted through the path 2, and their phase differences are calculated. This is shown as “S303′-S304 ′ phase difference” (path 2) in FIG.

  As shown in FIG. 2, the phase difference “S1′−S201 ′” output from the phase difference calculation unit 31 (referred to as phase difference φ1) and the phase difference “S102′−S2 ′” (referred to as phase difference φ2). Is input to the first filter 32. The phase difference “S301′-S302 ′” (referred to as phase difference φ3) and the phase difference “S303′-S304 ′” (referred to as phase difference φ4) output from the phase difference calculation unit 31 are the second values. Input to the filter 33.

  The superimposed component of each signal is removed by the trigonometric function calculator 43. However, there are cases where the superimposition component cannot be completely removed by the trigonometric function calculator 43 alone. For this purpose, a first filter 32 and a second filter 33 are provided. The first filter 32 and the second filter 33 also function as filters that remove the superimposed component.

  The first filter 32 removes the superimposed components of the signals of the phase difference φ1 (= S1′−S201 ′) and the phase difference φ2 (= S102′−S2 ′). As described above, when a superimposition component is generated in the signal, the natural frequency f0 and the reference signal frequency fa are different from each other, so that the superimposition component is superimposed as a vibration component.

  The phase difference φ1 and the phase difference φ2 are DC component signals, and the first filter 32 passes the DC components of the phase difference φ1 and the phase difference φ2. Thereby, the superimposed component is removed. Note that the first filter 32 may remove the superimposed component by averaging the phase difference φ1 and the phase difference φ2.

  The second filter 33 removes the superimposed components of the signals having the phase difference φ3 (= S301′−S302 ′) and the phase difference φ4 (= S303′−S304 ′). Also in this case, since the natural frequency f0 and the reference signal frequency fa are different from each other, the superimposed component is superimposed as a vibration component.

  Therefore, the second filter 33 passes the DC components of the phase difference φ3 and the phase difference φ4. Thereby, the superimposed component is removed. The second filter 33 may also remove the superimposed component by averaging the phase difference φ3 and the phase difference φ4.

  The circuit error calculation unit 34 calculates the circuit error of the path 1 using the phase difference φ3, and calculates the circuit error of the path 2 using the phase difference φ4. For the reference signals S301 to S304, a reference signal having a reference signal frequency fa different from the natural frequency f0 of the first vibration signal S1 and the second vibration signal S2 is used. Therefore, the phase difference φ3 or the phase difference φ4 does not simply indicate a circuit error.

  The circuit error calculator 34 receives the natural frequency f0 from the frequency calculator 36 and the reference signal frequency fa from the reference signal frequency selector 37. Therefore, an operation for approximating the circuit error is performed using the phase differences φ3 and φ4, the natural frequency f0, and the reference signal frequency fa.

  The phase difference φ3 indicates a circuit error of the path 1, and the phase difference φ4 indicates a circuit error of the path 2. The circuit error calculation unit 34 calculates the circuit error of the path 1 using the phase difference φ3, and outputs the path 1 circuit error (denoted as Δφ1). Further, the circuit error calculation unit 34 calculates the circuit error of the path 2 (denoted as Δφ2) using the phase difference φ4, and outputs the path 2 circuit error. The path 1 circuit error Δφ1 and the path 2 circuit error Δφ2 are input to the circuit error correction unit 35.

  The circuit error correction unit 35 corrects the phase difference φ1 or φ2 that has passed through the first filter 32 by using the path 1 circuit error Δφ1 or the path 2 circuit error Δφ2. The route 1 and the route 2 shown in FIG. 1 are different from each other, and each route is affected differently due to a difference in environmental conditions or the like, thereby causing a phase shift. This phase shift becomes a circuit error in the path. This circuit error is also called a zero compensation value. Accordingly, the circuit error correction unit 35 corrects the circuit error.

  The phase difference φ1 and the phase difference φ2 corrected by the circuit error correction unit 35 for the circuit error indicate the flow rate flowing through the pipeline. Therefore, the digital signal processing circuit 20 outputs the phase difference φ1 and the phase difference φ2 corrected for the circuit error to the CPU 21 as flow signals.

  The frequency calculation unit 36 calculates a frequency. For this purpose, as shown in FIG. 2, the first vibration signal S1 ′ is input from CH1 and the first vibration signal S102 ′ is input from CH3. The frequency of these signals is the natural frequency f0, and the frequency calculator 36 calculates the natural frequency f0 based on the first vibration signal S1 ′ or the first vibration signal S102 ′.

  The frequency of the second vibration signal S2 (and the second vibration signal S201) is also the natural frequency f0. Therefore, these signals may be input to the frequency calculation unit 36 to calculate the natural frequency f0.

  The natural frequency f0 calculated by the frequency calculation unit 36 is input to the circuit error calculation unit 34 and also to the reference signal frequency selection unit 37. The reference signal frequency selection unit 37 selects a frequency different from the natural frequency f0. This selected frequency is the reference signal frequency fa described above. The reference signal frequency fa is “fa = f0 + a”, which is a frequency different from the natural frequency f0 by an amount a.

  The reference signal frequency fa selected by the reference signal frequency selection unit 37 is output to the circuit error calculation unit 34 and also to the reference signal generation circuit 10. Accordingly, the reference signal generation circuit 10 outputs a reference signal having the reference signal frequency fa, and the reference signal is branched to become reference signals S301 to S304.

  The above is the configuration. Next, the operation will be described. The drive circuit 22 shown in FIG. 1 outputs the drive signal DRV, so that the drive coil L3 vibrates and vibrates the pipeline. The upstream coil L1 of the detector 2 detects vibration and outputs a first vibration signal S1. The downstream coil L2 detects vibration and outputs a second vibration signal S2.

  The RTD drive circuit 23 outputs a resistance drive signal RTDDRV. Thereby, the resistance temperature detector RTD measures the temperature of the fluid flowing through the pipe line. This measured temperature is output as a temperature signal TEMP. Therefore, the first vibration signal S1, the second vibration signal S2, and the temperature signal TEMP are input to the converter 3.

  As described above, path 1 and path 2 cause a path circuit error. In other words, the environmental conditions of the route change due to ambient temperature fluctuations, aging of circuits and wiring, and the like. This generates a circuit error in the path. When a circuit error occurs, a phase shift occurs and accurate flow measurement cannot be performed. For this reason, a circuit error is corrected.

  A circuit error is calculated in the path 1, and a phase difference is calculated in the path 2 in order to measure the flow rate in parallel. Thereby, it is possible to calculate the circuit error while calculating the phase difference for the flow rate measurement. For this reason, the flow measurement is not interrupted and the circuit error is not measured, and a wasteful time is reduced and a stable flow measurement is performed. The phase difference calculation and the circuit error calculation for the flow rate measurement are alternately performed in a time division manner.

  That is, in parallel with the calculation of the circuit error of the path 1, the phase difference calculation for the flow rate measurement is performed using the signal of the path 2. This is performed for a certain time, and in parallel with the calculation of the circuit error of the path 2, the phase difference calculation for the flow rate measurement is performed using the signal of the path 1. When this phase difference calculation is performed, an accurate flow rate measurement is performed by correcting the circuit error obtained last time. The above operation is repeated alternately in time division.

  First, the case where the circuit error of the path 1 is calculated and the phase difference calculation for the flow rate measurement is performed using the signal of the path 2 will be described. The first switching circuit 11 selects the reference signal S301. Therefore, the channel CH1 becomes the reference signal S301 ′, and the reference signal S301 ′ is amplified by the first input amplifier 15. The reference signal S301 ′ is converted into a digital signal by the phase difference ADC 19 and input to the digital signal processing circuit 20.

  The second switching circuit 12 selects the reference signal S302. Therefore, the channel CH2 becomes the reference signal S302 ′, and the reference signal S302 ′ is amplified by the second input amplifier 16. The reference signal S302 ′ is converted into a digital signal by the phase difference ADC 19 and input to the digital signal processing circuit 20.

  The third switching circuit 13 selects the first vibration signal S102. Accordingly, the channel CH3 becomes the first vibration signal S102 ′, and the first vibration signal S102 ′ is amplified by the third input amplifier 17. Then, the first vibration signal S <b> 102 ′ is converted into a digital signal by the phase difference ADC 19 and input to the digital signal processing circuit 20.

  The fourth switching circuit 14 selects the second vibration signal S2. Accordingly, the channel CH4 becomes the second vibration signal S2 ′, and the second vibration signal S2 ′ is amplified by the fourth input amplifier 18. Then, the second vibration signal S <b> 2 ′ is converted into a digital signal by the phase difference ADC 19 and input to the digital signal processing circuit 20.

  The operation of the digital signal processing circuit 20 will be described with reference to FIG. As described above, the channel CH1 is the reference signal S301 ′, the channel CH2 is the reference signal S302 ′, the channel CH3 is the first vibration signal S102 ′, and the channel CH4 is the second vibration signal S2 ′. These four signals are input to the phase difference calculation unit 31 shown in FIG.

  As described above, the signals transmitted through the four channels CH1 to CH4 may be superimposed on each other. When a superimposed component is added to each signal, an error occurs in various calculations performed using these signals. For example, an error occurs in the calculation of the phase difference calculation unit 31, and an error also occurs in the calculation of the circuit error calculation unit 34. As a result, an error occurs in the flow rate measurement, and the accuracy of the finally obtained flow rate value is lowered.

  The frequency of the reference signal S301 and the reference signal S302 (reference signal frequency fa) is set to a frequency different from the frequency (natural frequency f0) of the first vibration signal S1 and the second vibration signal S2. Therefore, although a superimposition component is generated in each signal, since the reference signal frequency fa and the natural frequency f0 are different from each other, the superimposition component is superimposed on each signal as a vibration component.

  This superposed component causes a measurement error. Therefore, the trigonometric function calculator 43 removes the superimposed component. As shown in FIG. 3, the reference signal S301 ′, the reference signal S302 ′, the first vibration signal S102 ′, and the second vibration signal S2 ′ are the first FIR filter 41 and the second FIR filter 42 that constitute the Hilbert transformer 40, respectively. Is input.

  The first FIR filter 41 outputs the input signal in the form of “A × sin (ωt)”, and the second FIR filter 42 outputs the input signal in the form of “A × cos (ωt)”. Then, the trigonometric function calculator 43 calculates the phase difference. The trigonometric function calculator 43 samples the phase difference by sampling at a plurality of times, and averages the phase difference at each time. Thereby, the vibration component superimposed on each signal is averaged. As a result, the superimposed component is removed from the signal.

  Here, the phase difference between the first vibration signal S102 ′ and the second vibration signal S2 ′ is calculated and output from the phase difference calculation unit 31. This phase difference φ2 indicates the flow rate of the fluid flowing through the pipeline, and a signal of the phase difference φ2 is input to the first filter 32. The signal of the phase difference φ2 when no superimposed component is generated is a DC signal.

  However, when the superimposed component is not sufficiently removed by the trigonometric function calculator 43, the superimposed component is superimposed on the signal having the phase difference φ2. As described above, since the natural frequency f0 and the reference signal frequency fa are different from each other, a superimposed component is superimposed as a vibration component on the DC signal component of the phase difference φ2.

  Therefore, the first filter 32 can remove the superimposed component by passing only the DC signal having the phase difference φ2. As a result, the signal having the phase difference φ2 subjected to the filtering action by the first filter 32 is input to the circuit error correction unit 35 in a state where the superimposed component is removed (that is, as a DC component signal).

  On the other hand, the trigonometric function calculator 43 calculates the phase difference between the reference signals S 303 ′ and S 304 ′, and the phase difference φ 3 is input to the second filter 33. The signal of the phase difference φ3 when no superimposed component is generated is a DC signal. However, also in this case, when the superimposed component is not sufficiently removed by the trigonometric function calculator 43, the superimposed component is superimposed on the signal having the phase difference φ3. This superimposed component is also superimposed as a vibration component.

  Therefore, the second filter 33 can remove the superimposed component by passing only the DC signal having the phase difference φ3. As a result, the signal of the phase difference φ3 that has been subjected to the filtering action by the first filter 32 is input to the circuit error calculation unit 34 in a state where the superimposed component is removed.

  The circuit error calculation unit 34 calculates the circuit error of the path 1 based on the phase difference φ3. The phase difference φ3 does not directly indicate the circuit error of the path 1. This is because the reference signals S303 ′ and S304 ′ for obtaining the phase difference φ3 have different reference signal frequencies fa from the natural frequency f0.

  The relationship between circuit error and frequency is not necessarily linearly proportional. Accordingly, since the reference signal frequency fa is different from the natural frequency f0, the phase difference φ3 does not directly indicate the circuit error of the path 1. FIG. 4 shows an example of the relationship between circuit error (unit: rad) and frequency (unit: Hz).

  Therefore, the circuit error calculation unit 34 shown in FIG. 2 performs an approximate calculation of the circuit error. For this purpose, the circuit error calculation unit 34 includes the phase difference φ3 from which the superimposed component has been removed by the second filter 33, the natural frequency f0 output from the frequency calculation unit 36, and the reference signal frequency selection unit 37 selected. The signal frequency fa is input.

  Here, since only one reference signal frequency fa is used as the reference signal, a first-order approximation calculation is performed. The circuit error of the path 1 is Δφ1 as described above. At this time, the circuit error Δφ1 of the path 1 is first-order approximated as shown in Equation 1 below.

  Thereby, the circuit error Δφ1 of the path 1 can be approximated and calculated. The circuit error Δφ1 of the path 1 is input to the circuit error calculation unit 34. As described above, the calculation of the circuit error Δφ1 of the path 1 and the calculation of the phase difference φ2 using the path 2 are performed in parallel. Next, the calculation of the circuit error Δφ2 of the path 2 and the calculation of the phase difference φ1 using the path 1 are performed in parallel.

  This operation is alternately performed in a time division manner. Therefore, the circuit error Δφ1 of the path 1 input to the circuit error correction unit 35 is corrected with respect to the phase difference φ1 calculated at the next time division timing. Thereby, the phase difference φ1 obtained by correcting the circuit error Δφ1 of the path 1 is obtained. That is, circuit error compensation (zero compensation) of path 1 is performed. And this phase difference (phi) 1 is output to CPU21 as a flow signal.

  The CPU 21 performs flow rate measurement based on the flow rate signal indicating the phase difference φ1. At this time, by performing correction using the temperature signal TEMP, it is possible to perform flow rate measurement that compensates for a measurement error due to the temperature of the fluid. This is output as flow rate data to the first output circuit 25, the second output circuit 26, and the display 27.

  As described above, the reference signal generation circuit 10 generates and outputs a reference signal whose reference signal frequency fa is different from the natural frequency f0 of the first vibration signal S1 and the second vibration signal S2. When a superimposed component of another signal occurs in the signals of each channel CH1 to CH4, a measurement error occurs. At this time, if the reference signal frequency fa and the natural frequency f0 are the same frequency, the superimposed component is a constant component (offset). Component), the superimposed component cannot be removed.

  However, in this embodiment, since the reference signal frequency fa and the natural frequency f0 are set to different frequencies, even if a superimposed component is generated in the signal, the superimposed component appears as a vibration component. Therefore, it is possible to remove the superimposition component by the filter processing, and it is possible to reduce the measurement error.

  In the above, the superimposed component is removed using the trigonometric function calculator 43, the first filter 32, and the second filter 33. That is, they serve as a filter that removes the superimposed component. However, the measurement error can be reduced without providing these filters.

  As described above, the reference signal frequency fa and the natural frequency f0 are set to different frequencies. Thereby, when a superimposition component arises in a signal, it superimposes as a vibration component instead of a constant component. If attention is paid to the center of the amplitude of the vibration component, a signal from which the vibration component is substantially removed can be obtained. Therefore, it is possible to reduce the influence of the superimposed component without providing a special filter.

  Further, if the averaging effect by the trigonometric function calculator 43 is sufficient, the superimposed component can be removed without using the first filter 32 and the second filter 33. In this case, the first filter 32 and the second filter 33 may not be provided.

  Further, the phase difference is calculated by the Hilbert transformer 40 and the trigonometric function calculator 43 shown in FIG. 3, and at the same time, the superimposed component is removed by the averaging effect. The phase difference may be calculated by another circuit other than the Hilbert transformer 40 and the trigonometric function calculator 43. In that case, it is desirable to provide the first filter 32 and the second filter 33.

  In this embodiment, the calculation of the circuit error Δφ1 of the path 1 and the calculation of the phase difference φ2 of the path 2 are performed in parallel, and then the calculation of the circuit error Δφ2 of the path 2 and the calculation of the phase difference φ1 of the path 1 are performed. However, the present invention is not limited to this example.

  The present embodiment can also be applied to a case where calculation of a circuit error and calculation of a phase difference for flow rate measurement are alternately performed. In this case, the circuit error calculation and the phase difference calculation can be performed using only one of the path 1 and the path 2 shown in FIG. Therefore, the circuit configuration is simple.

  However, by performing the calculation of the circuit error and the calculation of the phase difference for the flow rate measurement in parallel, there is an effect that it is possible to eliminate useless waiting time. Since this is necessary, signal superposition is likely to occur. Therefore, as in the present embodiment, by making the reference signal frequency fa different from the natural frequency f0, it is necessary to remove the superimposed component. Therefore, it is effective to apply this embodiment.

  The reference signal generation circuit 10 generates and outputs one type of reference signal (reference signal frequency fa) different from the natural frequency f0, but generates reference signals of two or more different frequencies. May be. For example, a reference signal having a reference signal frequency fa (= f0 + a) and a reference signal having a reference signal frequency fb (= f0 + b) may be output.

  When two or more types of reference signals having different frequencies are used, the accuracy of the circuit error calculation by the circuit error calculator 34 is improved. That is, when only the reference signal frequency fa is used, a first-order approximation calculation is performed using one point, but when using the reference signal frequencies fa and fb, a first-order approximation calculation using two points can be performed. .

  When two types of reference signal frequencies fa and fb are used, the reference signal generation circuit 10 alternately outputs a reference signal of the reference signal frequency fa and a reference signal of the reference signal frequency fb in a time division manner. The phase difference φ3 (= S303′−S304) obtained when the reference signal with the reference signal frequency fa is output is φ31, and the phase difference φ3 obtained when the reference signal with the reference signal frequency fb is output is φ32. Then, the circuit error Δφ1 of the path 1 is expressed by the following formula 2.

  In this case, since linear approximation by two points can be performed, the accuracy of approximation of the circuit error Δφ1 of the path 1 is improved as compared with the case where only the reference signal frequency fa is used. Therefore, the more accurate circuit error Δφ1 of the path 1 can be obtained, and the flow measurement error can be further reduced.

  Alternatively, the circuit error may be calculated using reference signals having three different reference signal frequencies. When three types of reference signals are used, a quadratic approximation calculation using three points is performed. In this case, a more accurate circuit error Δφ1 of the path 1 can be obtained. Of course, reference signals having four or more different reference signal frequencies may be used.

  The reference signal frequency fa is fa = f0 + a. The reference signal frequency fb is fb = f0 + b. At this time, a and b may be changed dynamically. However, the reference signal frequency may be fixedly set. FIG. 5 shows an example in which the reference signal frequencies fa and fb are fixedly held.

  In FIG. 5, a reference signal frequency holding unit 38 is added to the configuration of FIG. The reference signal frequency holding unit 38 holds the reference signal frequencies fa and fb fixedly. These have different frequencies from the natural frequency f0. Only the reference signal frequency fa may be held.

  The reference signal frequency selection unit 37 does not receive the reference signal from the frequency calculation unit 36 and receives the reference signal frequencies fa and fb from the reference signal frequency holding unit 38. The reference signal frequencies fa and fb are input to the circuit error calculation unit 34. As a result, the circuit error Δφ1 of the path 1 can be calculated.

  The reference signal frequencies fa and fb held by the reference signal frequency holding unit 38 may not be changed, or may be parameters that can be designated by the user. Further, the reference signal frequencies fa and fb may be selected according to the type and size of the detector 2.

  The reference signal frequency fa (same for fb) can be determined based on the approximation error when calculating the circuit error and the effect of removing the superimposed component by the filter. The reference signal frequency fa is fa = f0 + a, that is, a frequency that is different from the natural frequency f0 by a.

  At this time, if the difference between the reference signal frequency fa and the natural frequency f0 is too large, that is, if the value of a is too large, the approximation error of the circuit error calculation performed by the circuit error calculation unit 34 becomes large. In other words, if the difference between the reference signal frequency fa and the natural frequency f0 is small, that is, if the value of a is small, the approximation error can be reduced.

  On the other hand, if the difference between the reference signal frequency fa and the natural frequency f0 is too small, the reference signal frequency fa and the natural frequency f0 are close to each other, so that the superimposed component when the signal is superimposed becomes a constant component (offset component). Get closer. For this reason, even if various filters are used, it is difficult to remove the superimposed component. As a result, the measurement error cannot be reduced.

  Therefore, the selection of the reference signal frequency fa is set so that the approximation error when calculating the circuit error does not become large and the superimposed component when the signal is superimposed can be removed. Can be reduced.

  In FIG. 2, the first filter 32 is provided between the phase difference calculation unit 31 and the circuit error correction unit 35. However, as shown in FIG. You may arrange in. The second filter 33 needs to be arranged before the circuit error calculation unit 34. That is, the calculation accuracy is improved by calculating the circuit error using the phase difference signal from which the superimposed component is removed.

  On the other hand, the first filter 32 is a filter for removing a superimposed component from the phase difference signal between the first vibration signal S1 and the second vibration signal S2. Therefore, the superimposition component can also be removed from the phase difference signal by removing the superposition component after correction by the circuit error correction unit 35, and the measurement error can be reduced.

  FIG. 2 shows a configuration in which the first filter 32 and the second filter 33 are arranged after the phase difference calculation unit 31, but the first filter 32 and the second filter 33 are arranged before the phase difference calculation unit 31. It is good also as a structure to arrange.

DESCRIPTION OF SYMBOLS 1 Coriolis mass flowmeter 2 Detector 3 Converter 10 Reference signal generation circuits 11-14 Switching circuits 15-18 Input amplifier 19 Phase difference ADC
20 Digital signal processing circuit 21 CPU
22 Drive circuit 23 RTD drive circuit 24 Temperature ADC
31 phase difference calculation unit 32 first filter 33 second filter 34 circuit error calculation unit 35 circuit error correction unit 36 frequency calculation unit 37 reference signal frequency selection unit 38 reference signal frequency holding unit 40 Hilbert transformer 41 first FIR filter 42 second FIR Filter 43 Trigonometric function calculator S1 First vibration signal S2 Second vibration signal

Claims (6)

A Coriolis mass flowmeter that vibrates a pipeline through which fluid flows and measures the flow rate of fluid flowing through the pipeline based on a first vibration signal upstream and a second vibration signal downstream of the pipeline. And
A reference signal generation unit that generates a reference signal having a frequency different from the frequencies of the first vibration signal and the second vibration signal, which are natural frequencies of the pipe ,
The path of the first vibration signal, the second vibration signal, and the reference signal is branched into a plurality of phases, the phase difference between the first vibration signal and the second vibration signal, and the position between the branched reference signals. A phase difference calculation unit for calculating a phase difference;
A circuit error calculation unit that calculates a circuit error in the natural frequency by based rather approximate calculation of the phase difference between said reference signal,
A circuit error correction unit that corrects the circuit error with respect to the phase difference between the first vibration signal and the second vibration signal calculated by the phase difference calculation unit;
A flow rate calculation unit that calculates the flow rate based on a phase difference between the first vibration signal and the second vibration signal corrected by the circuit error correction unit;
A Coriolis mass flow meter characterized by comprising: The Coriolis mass flowmeter according to claim 1, further comprising a filter that removes a superimposed component when each signal transmitted through the branched path is superimposed on a signal transmitted through another path. The reference signal generation unit outputs two or more reference signals each having a different frequency in a time-sharing manner,
The Coriolis mass flowmeter according to claim 2, wherein the circuit error calculation unit calculates the circuit error by approximation of the first order or higher. The frequency of the reference signal is determined based on an approximation error when the circuit error is calculated by the circuit error calculation unit and an effect of removing the superimposed component by the filter. Coriolis mass flow meter. A reference signal frequency storage unit for storing a frequency different from the frequencies of the first vibration signal and the second vibration signal;
Select a frequency stored in the reference frequency storage unit, and output to the reference signal generation unit, a reference signal frequency selection unit,
The Coriolis mass flowmeter according to claim 1, comprising: 6. The calculation of the phase difference between the first vibration signal and the second vibration signal and the calculation of the phase difference between the reference signals are performed in parallel. 6. Coriolis mass flow meter as described.

Images