JPS5899717A - Ultrasonic wave flowmeter - Google Patents

Abstract

PURPOSE:To obtain the ultrasonic wave flowmeter without level detecting errors in received pulses due to dust and the like, by obtaining a proparating time by multiplying the measurement distribution characteristic of the propagating time of the ultrasonic wave and the weighting function which gives the weight, with the peak value of the distribution characteristic as a center. CONSTITUTION:A signal S1 is sent to a transmission circuit 13 from a synchronizing circuit 11 by a start signal S from a CPU 12, and a clock signal CL is concurrently sent to a counter 14 wherein an amount of delay is set, and down counting is performed. The signal S1 passes through a transmitter and a receiver 2a and 2b of a pipe 1. Then a signal S2 is outputted from a receiving circuit 15, and inputted to a high speed integrating circuit 16 as a stop signal, when the counted value 14 becomes a ''0''. Then, an integrated output dV1 is inputted to the CPU 12. The CPU 12 is constituted as follows: The weighting function of a ''1'' is formed in the region, wherein the dV1 that gives the maximum peak is the center, and that of ''0'' is formed in the other region; and a distribution curve is obtained by multiplying the result and the distributing characteristic. Therefore the level detecting errors in the signal S2 due to the dust and the like are eliminated.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an ultrasonic flowmeter that repeatedly transmits ultrasonic pulses and measures the flow rate by measuring the time from the transmission time of the ultrasonic pulses to the reception time of the ultrasonic pulses. .

The ultrasonic pulses used in ultrasonic flowmeters are ultrasonic carrier waves emitted in a pulsed manner, and the amplitude of the received ultrasonic pulses increases gradually from the time the reception starts, and then gradually attenuates. I'm going to go. In order to perform precise time measurement using such ultrasonic pulses, ultrasonic flowmeters, etc. detect specific waves (% radio waves) in the carrier wave of the received ultrasonic pulses, and The flow velocity and flow rate of the fluid to be measured are measured using the arrival time of a specific wave.

Fig. 1 is a diagram showing the flow rate measurement principle of this ultrasonic flowmeter. As shown in this figure, a pair of ultrasonic transducers (hereinafter simply referred to as transducers) is installed in a pipe 1 with an inner diameter of D. 2a and 2b are attached diagonally opposite to each other with respect to the axis of the conduit 1. The distance between these transducers 2&, 2b! , transducer 2
The angle between the line connecting m and 2b and the diameter of pipe 1 is θ, and the flow velocity of +11J constant fluid flowing in pipe 1 is! , if the propagation velocity of the ultrasonic wave in this fluid to be measured is C, then the propagation time (outward time) required for the ultrasonic pulse transmitted from the transducer 2& to be received by the transducer 2b is
T, is 'l' = l (1)
1-1Tηπr. Conversely, the propagation time (return time) T2 required for the ultrasonic pulse transmitted from the transducer 2b to be received by the transducer 2m is T2=+ (2). Also, is the distance l equal to D/a+Sθ?
), the following relationship is derived. Here, since the pipe diameter p and angle θ are constant, the flow velocity ν of the fluid to be measured can be found from equation (8) by measuring the propagation times Tx and T8 and finding the difference in their reciprocals. The flow rate can be obtained from this flow velocity υ and the pipe diameter.

Here, the waveform of the ultrasonic pulse received by the transducers 2& and 2b is as shown in 0) in Figure 2, and the upper envelope * (envelope) a of the carrier wave gradually changes from the start of reception. It rises and then gradually falls. On the other hand, a rectangular wave is generated in the receiving circuit of the ultrasonic flowmeter (see FIG. 2 (b)). Then, the receiving circuit uses the part (i% constant radio wave) where the ultrasonic pulse first exceeds the reference level L (in this example, the second wave).
, detect the time when this specific wave crosses level 0,
The received pulse S rising at this time (Fig. 2 G-e
). Then, the time when this received pulse S2 is emitted becomes the arrival time of the ultrasonic pulse.

By the way, if the received ultrasonic pulse is attenuated by dust, bubbles, etc. in the fluid, the waveform received by the transducer 2 & 2b becomes as shown in Fig. 2, for example.
In this ultrasonic pulse, the wave that first exceeds the reference level L is the 13th wave, and 1 from the specific wave (2nd wave) mentioned above.
This means that you are one wave behind the other. As a result, the output signal of the comparator is as shown in the same figure (e), and the received pulse 4 is (
). A time measurement error of time T, which corresponds to the time of approximately l waves, occurs between the received pulse S2 in this case and the received pulse S8 shown in FIG. 2 f-1. This time ΔT is the carrier frequency of the ultrasonic pulse −,
If 1) is set to 500 KHz, it becomes approximately 2 μs, which becomes a decisive error in an ultrasonic flow needle that requires time measurement to be performed in units of several n13.

Therefore, two methods described below have been conventionally used to solve this problem.

■ Ultrasonic propagation time obtained for each measurement? , (or T,) N times, and the resulting average propagation time TI! A method in which the new propagation time T is taken as a normal value only when the relationship: l Tl - Tlm l≦ΔT is established by comparing II (6 lm) with the newly obtained propagation time T1.

■ Count the number of times the ultrasonic pulse exceeds the reference level L for each measurement (two times in the case shown in Figure 2 (a)), and only when this count is the same as the value under normal conditions, calculate the new propagation time. How to import it as normal data.

However, in the method (2) described above, if there is a large amount of dust, bubbles, etc. in the fluid to be measured, the average propagation time T1
In some cases, m does not show the correct value, and in such cases, there is a problem that errors occur in the measurement results or the measurement operation becomes unstable.

In addition, in the method (2), when the envelope of the ultrasonic pulse (see a in FIG. 2, 0) changes, the data to be captured is frequently selected incorrectly.

In view of the above-mentioned 41fi, the present invention provides an ultrasonic flowmeter that can perform measurements stably and with high accuracy even when a large amount of dirt and bubbles are present in the fluid to be measured, and which allows measurement of propagation time to be distribution characteristic detection means for sampling and detecting the distribution characteristic of the propagation time; and weighting function setting means for setting a weighting function for weighting the distribution characteristic in a predetermined interval centered on the peak value of the distribution characteristic. and the flow rate is controlled based on the result of multiplying the distribution characteristic by the weighting function. .

Embodiments of the invention will be described below with reference to aspects 4.
-i1 First, the present invention will be explained in detail with reference to Figure 3 and Figure 4. In addition, in the following explanation, the propagation time T1
The explanation will be given using 0 measurement as an example, but the same applies to the case where the propagation time T5#1 is fixed.

FIG. 3(a) shows the distribution characteristics of the propagation time T when there is no dust, bubbles, etc. in the am constant fluid and the propagation of the ultrasonic pulse is good. As shown in this figure, the distribution characteristic waits for a single peak at the correct measured value Tpt. On the other hand, the third
Figure ←) shows the distribution characteristics when there is a lot of dust, bubbles, etc. in the fluid to be measured. T4
It has a peak at Each measurement value T giving this peak
P, ~'I'Pji are separated by Δτ, which will be clear from the explanation in FIG.

By the way, in this invention, even if the propagation of ultrasonic pulses is poor, instead of only adopting normal measurement values as in the past, we can also adopt abnormal light measurement values and obtain the results shown in Figure 3 (Ro). ), keep in mind the distribution characteristics shown in 2d interval, and in other intervals, the sadness function W that becomes OI.
(t), and this weighting function W (t,) and Figure 3 (
Multiply by the distribution characteristics shown in b), and from this result, the propagation time T
1.

Furthermore, the present invention takes the following measures to further improve measurement accuracy.

In general, an integrating circuit is used to measure the propagation time T1, which is the time from when an ultrasonic pulse is emitted to when a received pulse is emitted. As shown in (b), the predetermined open voltage (constant voltage) with a propagation time of '1'' is integrated.The voltage v1 obtained as a result corresponds to the propagation time T□. When the dynamic range is widened, the accuracy required of the integration circuit itself becomes stricter, so generally, the integration start time is delayed by a predetermined delay time τ (τ<T,) as shown in f9 of the same figure. Then, the voltage dv1 is obtained by integrating with a faster integration circuit.It goes without saying that the Ot pressure dl corresponds to the transmission time T1. In this invention, the latter method is adopted, and even slower a
The FRf value τ is appropriately changed depending on the temperature of the fluid, the inner diameter of the pipe line 1, etc. so that it becomes an optimal value. This is the delay time τ
If T□ is too small compared to the propagation time T□, the dynamic range of the integrating circuit must be widened, resulting in poor measurement accuracy.

Next, FIG. 5 is a block diagram showing the configuration of an embodiment of the present invention.

In this figure, 10 is a clock pulse generation circuit that generates the clock pulse OL shown in FIG. At the same time, it also supplies a clock pulse CTJ to the counter 14. The counter 14 has two operating modes: up-counting and down-counting, and switching between these modes is performed by a control signal supplied from the (!PU 12). Count until the reception pulse S2 is supplied from the reception circuit 15, and set the count value N to 0PU12.
The down-count operation is the operation of supplying the initial value X supplied from OFυ12 to the clock pulse O
It continues counting down with L, and when the count value reaches O, it is started by supplying a pulse signal to the high speed division circuit 16. The dual-speed integration circuit 16 is a circuit that integrates a predetermined constant voltage from the time when the pulse signal is supplied from the counter 14 until the reception pulse S2 is supplied. The output voltage dv1 of the erroneous integration circuit 16 is converted into a digital signal by the interface circuit 19 and outputted as C! It is supplied to PU12. 20
is an output circuit that creates signals to be supplied to indicating instruments, display devices, etc. based on signals supplied from the CPU 12. A memory 21 is composed of a ROM (read only memory) and a RAM (-) random access memory).The ROM stores programs used in the 0PU12, and the RAM stores various data. In addition, switches 17 and 18 are switches controlled by 0PU12.

(The control line from C! FU 12 is not shown.) When measuring the propagation time T1, the common contacts 17c and 18a touch the contacts 17&, 18&, respectively, and when measuring the propagation time T2, the common contacts 17c , 18o is the contact 171), 18
1) respectively. Note that 22 is a data bus.

Next, we will explain the operation of this *embodiment. This embodiment has mode 1, which sets the delay time τ mentioned above, and mode 2, which actually measures the propagation time T1. First, we will explain the operation of mode 1. explain.

When the power is turned on in the circuit shown in Fig. 5, the CPU
A control signal is supplied from the counter 12 to the counter 14, and the counter 14 enters an up-counting mode.

Next, a clock pulse OL shown in FIG. 6(&)K is supplied from the clock generation circuit 10 to the synchronization circuit 11. Then, a start signal S shown in FIG. 6(b) is sent from 0PU12.
is supplied to the synchronization circuit 11, the synchronization circuit 11 supplies a start pulse S synchronized with the clock pulse CxJ to the transmission circuit 13 as shown in FIG. 6(0), and
A clock/< pulse OL is supplied to the counter 14. From this time, the counter 14 starts counting the clock pulses o'XJ. Then, when the ultrasonic wave is transmitted from the transducer 3 & to the transducer 3b and the received wave is supplied to the receiving circuit 15, the receiving circuit 15 receives the received pulse S! shown in FIG.
supply to. At this point, the counting operation of the counter 14 is completed, and the counted value M (see FIG. 6 (&)) is transferred to the C+PU 12.
The data is stored in the memory 21 via the . At this time again, OPυ
Step 12 calculates the initial value X by performing the calculation of =N-1, and stores this initial value XO value in the memory 21. This mode,
1, the count value N stored in the memory 21 corresponds to the approximate value of the propagation time T1, as can be seen by comparing (&) and (Q in FIG. 6). X corresponds to the delay time τ as shown in FIG. 6(b).

Next, the operation in mode 2 will be explained.

In FIG. 5, 0Ptr12 outputs a control signal that puts the counter 14 in a down-count mode, and the counter 14
After setting the counter to the down count mode, the value K is supplied to the counter 14 at an initial stage. And clock generation circuit l
A clock pulse OL is supplied from O to the synchronous circuit 11,
Furthermore, the start signal S from the opU 12 is transmitted to the synchronous circuit 11.
When the clock pulse CL is supplied to the transmitting circuit 13, the synchronizing circuit 11 supplies the start pulse S□ shown in FIG. From this point on, the counter 14 counts down the initial value X using the clock pulse OL. On the other hand, when the ultrasonic wave is transmitted from the transducer 2a to the transducer 2b and the received wave is supplied to the receiving circuit 15, the receiving circuit 15 transmits the direct receiving pulse S2 shown in FIG. 6(d) to the high-speed integration circuit at time t2. 16. By the way, the counter 14 supplies a pulse signal to the high-speed integration circuit 21 when the count value reaches 0, but the time t at this time is slightly earlier than the time t2 at which the received pulse is also supplied to the high-speed integration circuit 16. . This is because the initial value is smaller than the count value N corresponding to the approximate value of the propagation time T1. Then, the fast integration circuit 16 performs an integration operation from time t2 to t2, and the resulting voltage dV.

is supplied to the interface circuit 19. The interface circuit 19 converts the voltage dv1 into a digital signal, and then stores it in the memory 21 via the 0PU12. The above is the measurement value acquisition operation in mode 2.

In this embodiment, it normally operates in this mode 2, and changes to the mode immediately after the power is turned on and at appropriate intervals, and the initial value
That is, the delay time τ is updated.

In the above explanation, the difference between the count value N and the initial value ) may be 8.

Now, in mode 2, the voltage a stored in the memory 21
It is clear from the above that the value of V has the distribution characteristics shown in FIG. 7 (0) when there are many bubbles, dust, etc. in the fluid to be measured. In this figure, ΔV is a voltage corresponding to ΔT shown in FIG. 3 (61ff). Here, 0PU12 is the measured value that gives the highest peak value (section 2 centered on l
The weight function f is January in d and roJ in other intervals.
Create it(V). Then, (!PU12 multiplies the weighting function W(V) and the distribution characteristic to obtain the distribution curve shown in e9 of Fig. can be easily understood from the figure.Then, apUi2 calculates the average value Vmeanl of this distribution curve.Also, in this mode 2, T!rIean2 corresponding to the transmission WJ time T2 is calculated by the same operation. Of course you can.

Here, 0PU12 squares the count value N stored in the memory 21, and also calculates Vmeanl from VIT+ean2.
and divide the subtraction result by (N2).

By the way, the fact that the count value N is proportional to the propagation time T1 is the sixth
It is clear from the figure, and therefore the following equation holds true.

N (X:T,...
(4) The propagation times T1 and T2 are generally very close values, and the difference therebetween is on the order of nl. Therefore (4
) can also be used as an expression.

Therefore, the following equation holds true for the calculation performed in '% optT12.

The right side of this equation (6) is equal to the left side of the above-mentioned equation (8).

Therefore, it can be seen that the calculation performed by apU12 is proportional to the flow velocity υ.

In this embodiment, when statistically processing the values of the voltage avl stored in the memory 21, batch processing may be used in which M values of the voltage dv1 are fetched and then processed, or the oldest value is processed for each measurement. - A moving process may be used in which the measured value is replaced with the latest value.

Further, intermediate processing between these may be used.

As shown in the figure ←) ~), the weighting functions W, (V) ~ Wn (7) corresponding to each peak of the distribution characteristics shown in the same figure (←) ~
Set. This weighting function W, (V) ~ Wn ('V) is calculated as shown in the following equation, that is, the measurement value (lVp□ ~ 6 V p H) that gives the valley peak is centered, and is opened in an interval of 2 (L), Set it to "0" in other wards.

Then, “the subfunction W set in this way, (V
)-W, (V) are respectively multiplied by the distribution characteristics shown in FIG. 80). As a result of this, Figures 8 (to) to (su) respectively.
We obtain the distribution curve shown in
Δv6 or other than those corresponding to 匝6 V p i,
By appropriately shifting in integer multiples of ΔV, the measured value d V p 1
superimpose it on the distribution curve corresponding to . The resulting distribution curve is shown in Figure 8). Then, the average value Vmeanl of the distribution curve shown in Figure 8) is calculated, and the calculation shown in equation (6) is performed using this average value -Vmeanl (and "Vmeanl").

In addition, the weighting function W (V) in this embodiment may be rectangular, or a function with a smooth curve may be used near the boundary where the function value changes, as shown in FIG. 9, for example. .

As explained above, according to the present invention, the delay time is set slightly smaller than the propagation time, and the delay time is reduced corresponding to the propagation time, and
Detecting the distribution characteristic of the measured value obtained from the difference between the propagation time and the delay time, and setting a weighting function that weights the distribution characteristic in a predetermined interval centered on the peak value of this distribution characteristic. Since the flow rate is calculated based on the result obtained from the product of the weighting function and the distribution characteristics,
Even if there are multiple dusts and bubbles in the fluid to be measured,
Highly accurate and highly stable measurements can be performed. Further, even when there is a temperature change, there is an advantage that the measured fil1 degree does not decrease.

[Brief explanation of drawings]

Figure 1 shows the measurement principle of the ultrasonic flow meter, Figure 20)
, -(f) is a diagram explaining the process in which the received pulse 8□ is emitted, and Figure 3 (a) is a diagram showing the distribution characteristics of the propagation time T□ when the fluid to be measured has good ultrasonic propagation properties. Figure 3←) shows the propagation time T/ when the ultrasonic propagation characteristics of the measured fluid are poor.
) A diagram showing the distribution characteristics, FIG. 4 is a diagram for explaining the delay time τ, FIG. 5 is a block diagram showing the configuration of an embodiment of the present invention, and FIGS. Waveform diagrams showing waveforms in each part of the circuit shown in FIG. 5, and FIG.
Figure 8 (0) to 8) is a diagram explaining the method of calculating l, Figure 8 is a diagram explaining another method of calculating the weighting function W (m and the average value Vmean l, Figure 9 is a diagram explaining the weighting function W(V)). It is a diagram showing other functional forms. 10...Clock generation circuit (approximate value measuring means),
11... Synchronous circuit (approximate value measuring means), 12...
...CPU (central processing unit; approximate value measuring means, delay time setting means 1 distribution characteristic detecting means, weighting function setting means, multiplication means), 21...Memory (approximate value measuring means, delay time setting means) , Bonfu characteristic detection means, weight function setting means). Figure 3 Figure 4

Claims (1)

[Claims] In an ultrasonic flow meter, the flow rate is measured by setting a delay time slightly smaller than the propagation time of the ultrasonic waves between two transducers and measuring the difference between this delay time and the propagation time. an approximate value measuring means for measuring an approximate value of the propagation time at each time; a delay time setting means for setting the delay time based on the measurement result of the approximate value measuring means; and a difference between the propagation time and the delay time. a distribution characteristic detection stage for detecting a distribution characteristic of measured values obtained from a distribution characteristic; a weighting function setting means for weighting the distribution characteristic in a predetermined interval centered on a peak value in the distribution characteristic; An ultrasonic flowmeter comprising: a multiplier for multiplying the distribution characteristic.