JPH04339217A - Electromagnetic flowmeter - Google Patents

Abstract

PURPOSE:To enable the computing of a flow rate with a range of cycle of a high frequency by computing a flow rate signal with a cycle of high frequency as base using a time constant, a low frequency and a high frequency amplitude of a differentiatial noise and a sampling data. CONSTITUTION:In a microprocessor (MPU) 19, a timing signal Ts2 is applied to an excitation circuit 15 and an excitation current If containing two kinds of frequencies... high an low... is applied to a measuring fluid Q as magnetic field through an excitation coil 14. When on a symmetry exists on the up stream and down stream sides with respect to a line aligning detection electrodes 11 and 12, low and high frequency differential noises are generated with respect to a noise electromotive force which is generated in a fluid Q between the electrodes 11 and 12. A voltage between the electrodes 11 and 12 is read into a memory of the MPU 19 through a differential amplifier 16, a sampling circuit 17 and an A/D converter 18. The MPU 19 uses a time constant, low and high frequency amplitudes of a differentiatial noise and a sampling data to compute a flow rate signal with the cycle of the high frequency.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an electromagnetic flowmeter that is excited at a high frequency, and more particularly to an electromagnetic flowmeter that is improved so that it is not affected by differential noise generated in a fluid to be measured.

0002

[Prior Art] The present applicant filed Japanese Patent Application No. 3-7316 (title of invention: Electromagnetic Flowmeter) in 1991 as an electromagnetic flowmeter of this type.
The application was filed on January 24th. In this electromagnetic flowmeter, a high-frequency current and a low-frequency current with a lower frequency are applied to the fluid to be measured by superimposing a high-frequency current and a low-frequency current with a lower frequency into an excitation coil using an excitation means, and then the high-frequency current is applied to the fluid to be measured using a sampling means. The voltage generated in the fluid to be measured is sampled according to the high frequency timing of the signal and output as sampling data, and the noise calculation means is used to calculate the time-series changes in this sampling data within the period of the low frequency of the low frequency current. A noise component unrelated to the flow rate of the measured fluid is calculated using a flow rate calculation means, and then a flow rate signal that is not affected by noise is calculated using this noise component and sampling data by a flow rate calculation means. It is.

0003

[Problems to be Solved by the Invention] However, in order to calculate the differential noise parameter, this conventional electromagnetic flowmeter first calculates a flow rate that does not change on the order of a low frequency period. However, there is a problem in that it can only be used in applications where one of the following conditions is met: there is no change in the DC noise of the electrode on the order of a low frequency cycle. Specifically, the former cannot be used for applications that require high-speed response, and the latter cannot be used for applications that measure fluids containing slurry, where the electrode potential becomes unstable.

0004

[Means for Solving the Problems] In order to solve the above problems, the present invention provides an excitation coil with a high frequency current that takes three steady values and a low frequency current that takes two steady values at a lower frequency. excitation means for applying a composite magnetic field generated by these currents to the flow measurement fluid by superimposing them; and sampling means for sampling the voltage generated at the electrode at each of the above steady-state values and outputting it as sampling data; Using the four previous sampling data sampled within at least one period of high frequency, these algebraic operations are used to remove high frequency components among the flow rate signal component, DC noise component, and differential noise generated at the electrode. a noise calculating means for calculating only the frequency differential voltage; a time constant calculating means for calculating the ratio of a plurality of low frequency differential voltages sampled within the same half period of the low frequency and calculating a time constant of the differential noise from this; low frequency amplitude calculation means for calculating the low frequency amplitude of the low frequency component of the differential noise using the time constant and the previous low frequency differential voltage; A high frequency amplitude calculating means for calculating the high frequency amplitude of the high frequency component of Flow rate calculation means for calculating .

[0005]

[Operation] The excitation means superimposes a high-frequency current that takes a three-value steady value in the excitation coil and a low-frequency current that takes a two-value steady value of a lower frequency, and generates the fluid to be measured by these currents. Apply a compound magnetic field. The sampling means samples the voltage generated at the electrode at each of the above steady-state values and outputs it as sampling data. The noise calculation means calculates the high frequency component of the flow rate signal component, DC noise component, and differential noise generated in the electrode by performing these algebraic calculations using the four previous sampling data sampled within at least one period of high frequency. is removed and only the low frequency differential voltage is calculated. Then, the time constant calculation means calculates the ratio of a plurality of low frequency differential voltages sampled within the same half period of the low frequency, calculates the time constant of the differential noise from this, and the low frequency amplitude calculation means calculates the time constant of the differential noise. The low frequency amplitude of the low frequency component of the differential noise is calculated using the low frequency differential voltage of . Regarding the high frequency component, the high frequency amplitude calculation means similarly calculates the high frequency amplitude of the high frequency component of the differential noise using the above time constant and the above low frequency amplitude. The flow rate calculating means calculates a flow rate signal based on the period of the high frequency using the above time constant, the above low frequency amplitude, the above high frequency amplitude, and the above sampling data.

[0006]

[Embodiments] Hereinafter, embodiments of the present invention will be explained with reference to the drawings. FIG. 1 is a block diagram showing the structure of one embodiment of the present invention. 10 is a non-magnetic conduit made of ceramic pipe or stainless steel whose inner surface is lined for insulation. Furthermore, a pair of detection electrodes 11 and 12 are fixed insulated from the conduit 10. Also, the measuring fluid Q
A wetted electrode 13 is connected to a common potential point COM for grounding.

[0007] An excitation coil 14 is fixed outside the conduit 10, and an excitation current If is supplied thereto as a constant current from a constant current source inside an excitation circuit 15. On the other hand, the detection electrodes 11 and 12 are connected to each input terminal of a differential amplifier 16 (for simplicity, the amplification degree is assumed to be 1), and after common mode noise is removed, the sampling circuit 17 is output to.

The sampling circuit 17 is composed of a sampling switch SW, a hold capacitor C, a buffer amplifier Q, etc., and samples the output voltage of the differential amplifier 16 using a sampling signal SMPH and holds it in the hold capacitor C. . Hold voltage V
H is taken into the microprocessor 19 via the analog/digital conversion circuit 18. microprocessor 1
9 executes a predetermined calculation using the captured data and outputs it to the output terminal 20 as a flow rate signal FQ, and also outputs a predetermined timing signal TS1 to the analog/digital conversion circuit 18 for analog/digital conversion. control.

The microprocessor 19 also outputs a sampling signal SMPH for controlling the sampling switch SW of the sampling circuit 17 to the sampling switch SW. In addition, the microprocessor 19 sends a timing signal TS2 to the excitation circuit 15 to switch the excitation current If.
The waveform of the excitation current If is controlled by outputting and switching the constant current. This excitation current If is supplied from the excitation circuit 15 to the excitation coil 14 as a composite current each having a high frequency component with a period of T and a low frequency component with a period of 2nT. Therefore, a magnetic field having a high frequency component and a low frequency component is applied to the measurement fluid Q, and the detection electrode 11
, 12, a voltage E containing these frequency components is generated.

Next, regarding the operation of the embodiment configured as described above, the differential noise explanation diagram shown in FIG. 2, the flowchart diagram shown in FIG. 3, the waveform diagram shown in FIG. 4, and the waveform diagram shown in FIG. This will be explained using an explanatory diagram and a calculation diagram shown in FIG. 6, but first the noise generated between the electrodes will be explained using a differential noise explanatory diagram shown in FIG.

The inside of the conduit 10 including the electrodes 11 and 12 is shown in FIG.
It is as shown in (a). Here, the noise electromotive force on the upstream side is e1, the downstream side is e2, the fluid impedance on the upstream side is R11, R12, the fluid impedance on the downstream side is R21, R22, and the electrode capacitance on the upstream side is C11, C.
12. The downstream electrode capacitances are C21 and C22.

FIG. 2(b) shows an arrangement of these components as an equivalent circuit. Each constant in this case is
C1 = (C11・C12)/(C11+C12)
(1
) C2 = (C21・C22)/(C21+C22
)
(2) R1 = R11 + R12

(3) R2 = R21 + R2
2
(4) becomes. Therefore, the time constant G of this circuit is G=C1
C2 (R1 +R2)/(C1 +C2)
It can be calculated as (5).

In the equivalent circuit shown in FIG. 2(b), electrode 1
The sum of the noise electromotive forces e1 and e2 generated in the measured fluid on the upstream and downstream sides of the line connecting 1 and 12 is defined as en0, and the asymmetry ( When the degree of asymmetry β) occurs, differential noise en (t) expressed by the following equation is generated between the electrodes 11 and 12. This differential noise en (t) can be calculated as shown in the following equation by performing Laplace transform and Laplace inverse transform with reference to FIG. 2(b).

[0014]

[Math 1]

[0015] However, K1 = [(βC2 - C1)/(1+β)
(C1 +C2)] −[(βR1
−R2 )/(1+β)(R1 +R2 )]
(7) K3 = (βR1 - R2)
/(1+β)(R1 +R2)
(8) Here, A is an initial value determined to satisfy the continuity condition of differential noise in each time domain.

The first term in equation (6) is a value generated in proportion to the noise electromotive force en0 generated in the measurement fluid, and the second term includes the integral of this noise electromotive force en0. Therefore, the first term can be removed by sampling at the sampling timing when the noise electromotive force en0 is zero, but the second and subsequent terms cannot be removed because they include integration. Therefore, the following explanation will be given by ignoring the first term and assuming that signal sampling is executed when the excitation current is in a steady state and the noise electromotive force is zero.

The microprocessor 19 supplies the timing signal Ts2 to the excitation circuit 15 to generate an excitation current If containing two types of frequencies, low frequency and high frequency, as shown in FIG. 4(d).
is applied to the excitation coil 14. Therefore, this exciting current If
contains the high frequency components and low frequency components shown in FIGS. 4(e) and 4(f), which are applied to the measurement fluid as a magnetic field. Therefore, in the fluid to be measured, there is a signal voltage Vs having a waveform similar to that shown in FIG. 4(d) and a noise electromotive force e.
n0 is generated, both of which include high frequency components and low frequency components shown in FIGS. 4(e) and 4(f). The noise electromotive force en0 generated in this measurement fluid has a high frequency component en0H shown in Fig. 4 (g) and a low frequency component en0 shown in Fig. 4 (h).
It can be considered separately into two parts, L.

If asymmetry exists between the upstream and downstream sides of the line connecting the electrodes 11 and 12, as can be seen from the equivalent circuit shown in FIG. 2(b), this equivalent circuit has a predetermined time constant. , as described later, between the electrodes there is a high frequency component en0H (
Corresponding to the high-frequency differential noise enH shown in Fig. 4(i) and the low-frequency component en0L (Fig. 4(h)), Frequency differential noise en L is generated.

For explanation of the following calculations, the symbols assigned to each part of the high frequency component of the excitation current shown in FIG. 4(d) are shown in FIG. 5(a).
The symbols assigned to each part of the low frequency component are shown in FIG. 5(b), and for the sake of simplicity, the rise and fall of the excitation current If is changed at a known constant rate tr as shown in FIG. The following explanation assumes that the Then, constant noise electromotive force +En, -En will occur only during the period when the excitation current is changing at a constant rate tr (Fig. 4 (g
), (h)).

Therefore, these noise electromotive force en0H
, en0L into equation (6), and using the continuous condition for the high frequency component and the condition that the noise electromotive force is reversed after the differential noise is sufficiently attenuated for the low frequency component, Differential noise en H
, en L is calculated using the following equations (9) to (16)
, formulas (17) to (33) are obtained. The differential noises en H and en L at this time are shown in FIGS. 4(i) and (j).

First, the differential noise en H of the high frequency component
If T is the period of the high frequency, 0≦t≦te1,
en H = En K1 (1-e-t/G)
+A0 e-t/G
(9) When te1≦t≦T/4, en H =A1 e-(t-te1)/G

(10) If T/4≦t≦te2, en H = −En K1 (1−e−(t−
T/4)/G)+A2 e-(t-T/4)/G (
11) If te2≦t≦T/2, en H =A3 e-(t-te2)/G

(12) If T/2≦t≦te3, en H = −En K1 (1−e−(t−
T/2)/G)+A4 e-(t-T/2)/G (
13) If te3≦t≦3T/4, en H =A5 e-(t-te3)/G

(14) 3T/4≦t≦te4, en H =En K1 (1-e-(t-3T/
4)/G )+A6 e-(t-3T/4)/G
(15) For te4 ≦t≦T, en H =A7 e-(t-te4)/G

(16).

Here, in equations (9) to (16), t=0(T
), te1, T/4, te2, T/2, te3,
Substituting the continuity condition at 3T/4 and te4, the constant A
0 to A7 can be determined as follows. A0 = En K1 (1-e-t1/G)
・(1-e-T/4/G)e-t2/G
/(1+e-T/2/G)

(17) A1 = En K1 (1-e-t1
/G)・(1+e-T/4/G)/(1+e-T/2
/G) =AH (1+e-T/4/G
)
(18) However, AH = En K1 (1-e-t1
/G)/(1+e-T/2/G) (19
). A2 = En K1 (1-e-t1/G)
・(1+e-T/4/G)e-t2/G
/(1+e-T/2/G)

(20) A3 =-En K1 (1-e-t
1/G )・(1-e-T/4/G)/(1+e-T/
2/G) =-AH (1-e-T/4
/G)
(21) A4 =-En K
1 (1-e-t1/G)・(1-e-T/4/G)
e-t2/G /(1+e-T/
2/G)
(22) A5 =
-En K1 (1-e-t1/G)・(1+e-T
/4/G)/(1+e-T/2/G)
=-AH (1+e-T/4/G)
(23)
A6 =-En K1 (1-e-t1/G
)・(1+e-T/4/G)e-t2/G
/(1+e-T/2/G)

(24) A7 = En K1 (1-e-t
1/G )・(1-e-T/4/G)/(1+e-T/
2/G) =AH (1-e-T/4/
G)
(25)

Next, the low-frequency differential noise en L is expressed as en L =En K1 (1-e-t/G) when 0≦t≦te1.
+B0 e-t/G
(26) If te1 ≦t≦nT, en L =B1 e-(t-te1)/G

(27) If nT≦t≦nT+te1, en L =−En K1 (1−e−(t−
nT)/G)+B2 e-(t-nT)/G)(2
8) When nT+te1 ≦t≦2nT, en L =B3 e-(t-nT-te1)
/G)
(29)

Here, (26)
~(29), t=0(2nT), te1, nT, n
Substituting the continuity condition at T+te1, the constant B0 ~
B3 is B0 = -En K1 (1-e-
t1/G )e-(nT-t1)/G(1+e-nT/
G)

(30) B1
=En K1 (1-e-t1/G) (1+e-n
T/G)=AL

(31) B2
=En K1 (1-e-t1/G)e-(nT-
t1)/G(1+e-nT/G)


(32) B3 =-En K1 (1-
e-t1/G) (1+e-nT/G)
=-AL

It can be obtained as (33). Although the latter half period of the differential noise en L of the low frequency component is omitted here, it can be obtained in the same manner. These arithmetic expressions calculated based on the equivalent circuit shown in FIG. 2(b) are stored in advance in the memory of the microprocessor 19.

First, step 1 of the calculation flow shown in FIG.
I will explain about it. Now, as shown in FIG. 5(a), the excitation current If reaches a steady value at timings T/4 and T/2.
, 3T/4, T, ..., the voltage between the electrodes 11 and 12 is sampled by the sampling circuit 17 (FIG. 4(c)) and read into the memory of the microprocessor 19 via the analog/digital conversion circuit 18 (FIG. 4(c)). 3. Step 1) and
As shown in the calculation diagram of FIG. 6, the sampling voltages e01, e02, e03, e04 (calculation samples No. 1 to 4) stored in this memory..., eK1, eK2, eK3,
eK4 (calculation samples No. 7 to 10)..., e2K1
, e2K2, e2K3, e2K4 (calculation samples No. 12 to 15)... respectively have a column for flow rate signal eS, a column for electrode DC noise eD, a column for differential noise of high frequency components enH, and a column for differential noise of low frequency components. en
The signal listed in the L column and each noise are superimposed.

Next, the process moves to step 2 of calculating low frequency noise in the flow shown in FIG. Four sampling voltages (e01, e02, e03, e
04), (eK1, eK2, eK3, eK4), (e2
K1 , e2K2 , e2K3 , e2K4 ) are used to calculate calculation sample NO. By executing the algebraic calculation program described in the calculation formula columns of 5, 11, and 16,
The flow rate signal eS, the electrode DC noise eD, and the high frequency component enH are removed to calculate the low frequency noises e0, eK, and e2K, which are composed only of the low frequency component enL shown in the calculation diagram of FIG. 6, respectively. .

Next, the process moves to step 3 in which the time constant G of the differential noise (en H , en L ) in the calculation flow shown in FIG. 3 is calculated. For this purpose, calculation sample No. shown in the calculation diagram of FIG. By calculating the ratio (e0/eK) of the low frequency noise e0 and eK calculated in 5 and 11, the following (34) can be obtained. A program for this ratio calculation is also stored in the memory of the microprocessor 19 in advance.

[Math 2]

Next, the process moves to step 4 in which the amplitude AL of the low frequency component en L of the differential noise in the calculation flow shown in FIG. 3 is calculated. For this purpose, the time constant G determined by equation (34) must be set to the calculation sample No. 6 in FIG. The amplitude AL can be obtained by substituting it into the equation shown in 5. In addition,
The high frequency period T and t1 shown in FIG. 5 are known values. The program for this calculation is also programmed in advance by the microprocessor 19.
is stored in memory.

Similarly, the process moves to step 5 in which the amplitude AH of the high frequency component en H of the differential noise in the flow shown in FIG. 3 is calculated. From equation (19) and (31), (AH/AL)=(1+e-nT/G)
/(1+e-T/2/G) (35) is obtained. Since G, n, and T are known values, the amplitude AH of the high frequency component en H of the differential noise is expressed as (35)
It can be obtained from the formula. The program for this calculation is also stored in the memory of the microprocessor 19 in advance.

Next, the process moves to step 6 in which the calculation for removing the electrode DC noise eD in the flow shown in FIG. 3 is executed. This calculation is performed using the calculation sample No. shown in the calculation diagram of FIG. This is executed using the calculation formula shown in 6. The calculation result e0' obtained by this calculation includes a flow rate signal eS as shown in FIG.
is included, but electrode DC noise eD is removed. Similarly, although not shown in FIG. 6, the same algebraic operation is performed on the sampling voltages
eK4), (e2K1, e2K2, e2K3, e
2K4) etc. to reduce electrode DC noise eD.
is removed, but the flow rate signal eS is included.
, e2K', etc. can be obtained.

Next, the process moves to step 7 in which the flow rate signal eS in the flow shown in FIG. 3 is calculated. Operation sample No. 6 in FIG. The calculation result e0' calculated using the calculation formula shown in 6 includes a flow rate signal eS, a high frequency component e
n H , includes low frequency components en L , G, A
Since H and AL have been calculated as described above, from these, an unknown flow rate signal e0 S in the range of high frequency period T is obtained.
can be calculated. Similarly, the flow rate signals eK s , e2KS , ..., using the high frequency period T as a unit,
can be obtained for each.

Next, the process moves to step 8 in which the cycle for correcting high-frequency differential noise in the calculation flow shown in FIG. 3 is determined. Since the differential noise does not always vary, its parameters are corrected here in units of a half period of the low frequency, that is, nT. In this step 8, it is determined whether or not the period for performing this correction has arrived, and when this period has been reached, the calculations in steps 2 to 7 are executed again.

[0033] In addition to the above-mentioned calculations, there are various calculation means for calculating the flow rate in step 7. For example, the value e01 sampled at high frequency timing
, e02, e03, e04, etc. for the four consecutive values obtained by removing the differential noise en (
+, -, -, +) to calculate the electrode DC noise eD.
It is also possible to obtain a flow rate signal by removing the value e sampled at high frequency timing.
For the five consecutive values obtained by removing the differential noise en from 01, e02, e03, e04, e11, e12, etc., a sampling voltage of a part of the period T of the next high frequency is used (-1, +2, +0, -2, +1) to obtain the flow rate signal.

[0034]

Effects of the Invention As described above in detail with the embodiments, according to the present invention, the period of low frequency is
Since the flow rate signal can be calculated within a high frequency period range without requiring conditions such as no change in flow rate or no change in DC noise at the electrode, it is suitable for applications that require high-speed response or It can be used in applications that measure fluids containing slurries.

[Brief explanation of drawings]

FIG. 1 is a configuration diagram showing the configuration of one embodiment of the present invention.

FIG. 2 is an explanatory diagram illustrating the mechanism by which differential noise occurs.

FIG. 3 is a calculation flow diagram showing the calculation flow of the embodiment shown in FIG. 1;

FIG. 4 is a waveform diagram showing waveforms at various parts of the embodiment shown in FIG. 1;

FIG. 5 is a waveform explanatory diagram used to explain the waveform diagram shown in FIG. 4;

FIG. 6 is a calculation diagram illustrating details of each content of the calculation flow shown in FIG. 3;

[Explanation of symbols]

10 Conduit 11, 12 Detection electrode 14 Excitation coil 22 Excitation circuit 17 Sampling circuit 21 Microprocessor

Claims (1)

[Claims] Claim 1: A complex magnetic field generated by these currents in a fluid to be measured by superimposing a high-frequency current that takes a three-value steady value and a low-frequency current that takes a two-value steady value at a lower frequency in an excitation coil. excitation means for applying the voltage, sampling means for sampling the voltage generated in the electrode at each of the steady-state values and outputting the sampled data as sampling data, a noise calculating means for calculating only a low frequency differential voltage by removing high frequency components of the flow rate signal component, DC noise component and differential noise generated in the electrode by these algebraic operations; a time constant calculating means for calculating a time constant of the differential noise by calculating a ratio of the plurality of low frequency differential voltages sampled within the differential noise; a low frequency amplitude calculation means for calculating a low frequency amplitude of a frequency component; a high frequency amplitude calculation means for calculating a high frequency amplitude of a high frequency component of the differential noise using the time constant and the low frequency amplitude; An electromagnetic flowmeter comprising flow rate calculation means for calculating a flow rate signal based on a high frequency cycle using the low frequency amplitude, the high frequency amplitude, and the sampling data.