JP5512007B1 - Detection method using DFT cross-correlation method - Google Patents

Abstract

An ultrasonic flowmeter that is resistant to noise is provided.
An ultrasonic flowmeter includes a pair of ultrasonic elements that are arranged opposite to each other with a flow path through which a fluid to be measured flows, and that transmits and receives a predetermined ultrasonic signal. Fourier transform means for generating a third signal and a fourth signal by performing a Fourier transform on each of the first signal that is the received output of the ultrasonic element and the second signal that is the received output of the downstream ultrasonic element And a calculation means for generating a conjugate complex signal of one of the third signal and the fourth signal, calculating an inner product of the other signal and the conjugate complex signal, and generating a fifth signal; Inverse Fourier transform means for performing the inverse Fourier transform to generate the sixth signal, and calculation means for calculating the flow rate of the fluid to be measured based on the timing of the maximum value in the sixth signal.
[Selection] Figure 1

Description

  The present invention relates to an ultrasonic flow meter.

  As one form of an ultrasonic flow meter for measuring the flow rate of a fluid such as water, an ultrasonic flow meter having a pair of ultrasonic elements arranged opposite to each other with the flow channel between the upstream side and the downstream side of the flow channel through which the fluid flows It has been known.

  This ultrasonic flowmeter has a time tu until the ultrasonic signal transmitted from the upstream ultrasonic element reaches the downstream ultrasonic element and the ultrasonic signal transmitted from the downstream ultrasonic element. Based on the difference Δt between the time td to reach the upstream ultrasonic element and the difference Δt, the average flow velocity and flow rate of the fluid flowing through the flow path are obtained.

  As a method for obtaining the difference Δt between the time td and the time tu, a “zero cross method” and a “sum of products operation method” are known. The “product-sum operation method” belongs to the “cross-correlation method”.

  In the zero cross method (see Patent Document 1), an ultrasonic signal having periodicity is used, and the arrival time of the ultrasonic signal at each ultrasonic element is measured according to a specific condition for determining the arrival timing of the ultrasonic signal. The difference in arrival time is obtained by an analog method using hardware or a digital method using software.

  As an example of the specific condition, a condition “the timing at which the ultrasonic reception signal becomes zero voltage after first exceeding the threshold voltage (zero cross point) is set as the arrival timing” is used.

  On the other hand, in the product-sum operation method, the cross-correlation between the upstream reception signal and the downstream reception signal is calculated by the product-sum operation, and the difference in arrival time is obtained from the cross-correlation value.

  Specifically, in the product-sum operation method, the upstream side received signal and the downstream side received signal are arranged on the time axis, and the downstream side received signal on the time axis so that the upstream side received signal and the downstream side received signal are close to each other. Shift the signal step by step. Each time the downstream reception signal is shifted, the values of the upstream reception signal and the downstream reception signal at the same time on the time axis are multiplied at the same time, and the multiplied values are added, and the addition is performed. Is calculated as a cross-correlation value at the shift time. Then, the shift time that maximizes the cross-correlation value is obtained as the difference in arrival time.

JP 2012-193966 A

  In the zero cross method, when noise is applied to the zero cross point, the zero cross point moves correspondingly and the accuracy of the flow rate measurement deteriorates. That is, in the zero cross method, noise at a specific measurement point, which is a zero cross point, greatly affects the measurement accuracy.

  In the product-sum operation method, the magnitude of the received signal (received signal) itself is weighted. For this reason, the value obtained by multiplying the largest values in each received signal is weighted the most. Therefore, the noise level at the point where each received signal has the largest value greatly affects the measurement accuracy.

  Therefore, there is a problem that noise at a specific measurement point greatly affects the measurement accuracy in both the zero cross method and the product-sum operation method.

  The objective of this invention is providing the ultrasonic flowmeter which can solve the said problem.

The ultrasonic flowmeter of the present invention is
A pair of ultrasonic elements that are arranged opposite to each other across the flow path on the upstream side and downstream side of the flow path through which the fluid to be measured flows, and transmit and receive a predetermined ultrasonic signal;
For each of the first signal that is the reception output of the upstream ultrasonic element and the second signal that is the reception output of the downstream ultrasonic element, the third signal and the fourth signal are subjected to Fourier transform. Fourier transform means for generating
Arithmetic means for generating a conjugate complex signal of one of the third signal and the fourth signal, calculating an inner product of the other signal and the conjugate complex signal, and generating a fifth signal;
An inverse Fourier transform means for performing an inverse Fourier transform on the fifth signal to generate a sixth signal;
Based on the timing of the maximum value in said sixth signal, seen including a calculation unit, a for calculating the flow rate of the fluid to be measured,
The inverse Fourier transform means generates the sixth signal by performing the inverse Fourier transform on the fifth signal at a predetermined time interval,
The calculating means specifies the timing of the maximum value in the sixth signal, and includes a straight line having a derivative value of the sixth signal at a timing before the predetermined time as a slope from the timing of the maximum value, and the maximum value A straight line having a slope of a differential value of the sixth signal at the timing after the predetermined time from the timing, and the sixth signal between the timing before the predetermined time and the timing after the predetermined time is a straight line. Interpolation is performed, and the timing of the maximum value is calculated based on the result of the linear interpolation.

  According to the present invention, Fourier transform is used to calculate a cross-correlation function (sixth signal) between the first signal and the second signal. In the Fourier transform, calculation using all of the first signal and calculation using all of the second signal are performed. For this reason, the cross-correlation function calculated using the Fourier transform is less susceptible to noise at specific measurement points of the first signal and the second signal. Therefore, it is possible to suppress the noise at a specific measurement point from greatly affecting the measurement accuracy.

It is the figure which showed the ultrasonic flowmeter of one Embodiment of this invention. It is a flowchart for demonstrating operation | movement of an ultrasonic flowmeter. It is a flowchart for demonstrating operation | movement of an ultrasonic flowmeter. It is a figure for demonstrating the operation | movement which calculates the flow volume of to-be-measured fluid using the signal f (t) and the signal g (t). It is the figure which showed an example of the signal f (t) and the signal g (t). It is the figure which showed an example of signal F (s) and signal G (s). FIG. 6 is a diagram showing a real part of a signal c (t). It is the figure which showed the ultrasonic flowmeter of other embodiment of this invention.

  Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

  FIG. 1 is a diagram showing an ultrasonic flowmeter according to an embodiment of the present invention.

  In FIG. 1, an ultrasonic flowmeter includes ultrasonic elements 1 and 2, a transmission signal generator 3, a power amplifier 4, multiplexers 5 and 6, a preamplifier 7, an AGC (Auto Gain Controller) 8, and an AD. An (Analog-Digital) converter 9, a storage unit 10, a processing unit 11, and a display unit 12 are included.

  The processing unit 11 includes a measurement instruction unit 111, a Fourier transform unit 112, a calculation unit 113, an inverse Fourier transform unit 114, and a calculation unit 115 as functional blocks.

  The ultrasonic elements 1 and 2 are a pair of ultrasonic transducers disposed opposite to each other with the flow path 200 interposed between the upstream side and the downstream side of the flow path 200 through which a fluid to be measured (for example, a fluid such as water) flows. In the present embodiment, the ultrasonic element 1 is disposed on the upstream side of the flow path 200, and the ultrasonic element 2 is disposed on the downstream side of the flow path 200.

  The ultrasonic element 1 and the ultrasonic element 2 transmit and receive a 4 MHz ultrasonic signal through the fluid to be measured flowing through the flow path 200. The 4 MHz ultrasonic signal is an example of a predetermined ultrasonic signal and an ultrasonic signal having a predetermined frequency. In addition, the frequency of the ultrasonic signal transmitted / received is not limited to 4 MHz, and can be changed as appropriate. For example, a frequency of 1 to 4 MHz or a frequency of 8 MHz may be used.

  The transmission signal generator 3 generates a transmission signal for vibrating (driving) the ultrasonic element 1 and the ultrasonic element 2. In the present embodiment, each transmission signal is output for a predetermined output time T1.

  The power amplifier 4 amplifies the transmission signal generated by the transmission signal generator 3 so as to match the excitation voltage of the ultrasonic element 1 and the ultrasonic element 2.

  The multiplexer 5 switches the output destination of the transmission signal amplified by the power amplifier 4 between the ultrasonic element 1 and the ultrasonic element 2. The ultrasonic element 1 transmits a 4 MHz ultrasonic signal to the ultrasonic element 2 according to the input of the transmission signal, and the ultrasonic element 2 transmits the ultrasonic wave of 4 MHz according to the input of the transmission signal. The signal is transmitted toward the ultrasonic element 1. That is, the multiplexer 5 selects an ultrasonic element that is a transmission side.

  The multiplexer 6 receives the reception output (analog output) of the ultrasonic element 1 and the reception output (analog output) of the ultrasonic element 2, and receives the reception output of the ultrasonic element 1 and the reception output of the ultrasonic element 2. Alternatively, it is output to the preamplifier 7. That is, the multiplexer 6 selects an ultrasonic element that is a receiving side. The reception output of the ultrasonic element 1 is an example of a first signal, and the reception output of the ultrasonic element 2 is an example of a second signal.

  The preamplifier 7 amplifies the output of the multiplexer 6, that is, each reception output that is alternatively output (the reception output of the ultrasonic element 1 and the reception output of the ultrasonic element 2) to the input voltage range of the AGC 8.

  The AGC 8 amplifies each received output amplified by the preamplifier 7 according to the input voltage range of the AD converter 9. The intensity of each received output varies depending on the flow velocity of the fluid to be measured. The AGC 8 amplifies each reception output amplified by the preamplifier 7 in order to compensate for this variation.

  The AD converter 9 converts the output (each received output) of the AGC 8 that is an analog signal into a digital signal. Specifically, the AD converter 9 samples the output of the AGC 8 (each received output) at a constant time interval (sampling time) T2 and converts it into a digital signal (digitized data data). T2 is shorter than T1. In the AD converter 9, the maximum numerical value when digitizing is determined. For example, the 10-bit AD converter 9 converts an analog value into an integer value of −512 to 511.

  The storage unit 10 is, for example, a RAM (Random Access Memory), and stores digitized digital data (each digitized reception output).

  The processing unit 11 is, for example, a CPU (Central Processing Unit) or an MPU (Micro Processing Unit), and calculates the flow rate of the fluid to be measured flowing through the flow path 200 using the digital data stored in the storage unit 10.

  The measurement instructing unit 111 outputs a switching signal to the multiplexers 5 and 6 when measuring the flow rate. The multiplexers 5 and 6 are configured so that the ultrasonic element 1 is a transmission side (a side where a transmission signal is provided) and the ultrasonic element 2 is a reception side (a reception output is selected by the multiplexer 6) according to the switching signal. And a state in which the ultrasonic element 2 is the transmission side and the ultrasonic element 1 is the reception side. Further, the measurement instruction unit 111 outputs an operation signal to the transmission signal generator 3. The transmission signal generator 3 generates a transmission signal when receiving the operation signal.

  The Fourier transform unit 112 is an example of a Fourier transform unit. The Fourier transform unit 112 performs Fourier transform (for example, DFT (Discrete Fourier Transform)) for each digitized reception output in the storage unit 10.

  Hereinafter, among the digitized reception outputs in the storage unit 10, the reception output of the ultrasonic element 1 is referred to as “signal f (t)”, and the reception output of the ultrasonic element 2 is referred to as “signal g (t)”. Called. The result of Fourier transform of the signal f (t) is referred to as “signal F (s)”, and the result of Fourier transform of the signal g (t) is referred to as “signal G (s)”.

  The signal f (t) is also an example of the first signal, the signal g (t) is also an example of the second signal, the signal F (s) is an example of the third signal, and the signal G (s) is It is an example of a 4th signal.

  The calculation unit 113 is an example of a calculation unit. The calculation unit 113 generates a conjugate complex signal of one of the signals F (s) and G (s), calculates an inner product of the other signal and the conjugate complex signal, and calculates the signal f (t) and The cross spectrum of the signal g (t) is calculated.

  In the present embodiment, the calculation unit 113 generates a conjugate complex signal (hereinafter referred to as “signal H (s)”) of the signal G (s). The calculation unit 113 calculates the inner product (F (s) · H (s)) of the signal F (s) and the signal H (s). The calculation result of the inner product of the signal F (s) and the signal H (s) is an example of the fifth signal and represents the cross spectrum of the signal f (t) and the signal g (t).

  The inverse Fourier transform unit 114 is an example of an inverse Fourier transform unit. The inverse Fourier transform unit 114 performs inverse Fourier transform (for example, IDFT (Inverse Discrete Fourier Transform)) on the calculation result of the inner product of the signal F (s) and the signal H (s) to generate the signal c (t). . The signal c (t) is an example of a sixth signal and represents a cross-correlation function between the signal f (t) and the signal g (t).

  The calculation unit 115 is an example of a calculation unit. The calculation unit 115 calculates the flow rate of the fluid to be measured based on ta (timing) that gives the maximum value of the signal c (t).

  Next, the operation will be described.

  FIG. 2 is a flowchart for explaining the operation until the signal F (s) and the signal G (s) are generated among the operations of the ultrasonic flowmeter. FIG. 3 is a flowchart for explaining the operation of calculating the flow rate of the fluid to be measured using the signal F (s) and the signal G (s). FIG. 4 is a diagram for explaining the operation of calculating the flow rate of the fluid to be measured using the signal f (t) and the signal g (t).

  FIG. 5 is a diagram illustrating an example of the signal f (t) and the signal g (t). FIG. 6 is a diagram illustrating an example of the signal F (s) and the signal G (s), and FIG. 6A is a diagram illustrating the real part of each of the signal F (s) and the signal G (s). FIG. 6B is a diagram illustrating the imaginary part of each of the signal F (s) and the signal G (s). FIG. 7 is a diagram showing the real part of the signal c (t).

  At the time of measuring the flow rate, the measurement instruction unit 111 outputs to the multiplexers 5 and 6 a switching signal for selecting the ultrasonic element 2 by the multiplexer 5 and selecting the ultrasonic element 1 by the multiplexer 6 (step S201).

  Upon receiving the switching signal, the multiplexer 5 selects the ultrasonic element 2 as the transmission side (step S202). Further, when receiving the switching signal, the multiplexer 6 selects the ultrasonic element 1 as the receiving side (step S203).

  The measurement instruction unit 111 outputs an operation signal to the transmission signal generator 3 after a specific time T3 after outputting the switching signal in Step 201 (Step S204).

  When receiving the operation signal, the transmission signal generator 3 outputs the transmission signal for the output time T1 (step S205).

  The transmission signal is amplified by the power amplifier 4 so as to match the excitation voltage of the ultrasonic element 1 and the ultrasonic element 2, and then reaches the ultrasonic element 2 via the multiplexer 5. The ultrasonic element 2 transmits an ultrasonic signal of 4 MHz to the ultrasonic element 1 through the fluid to be measured in the flow path 200 for an output time T1 in accordance with the power-amplified transmission signal (step S1). S206).

  On the other hand, the reception output of the ultrasonic element 1 is output to the preamplifier 7 after the multiplexer 6 selects the ultrasonic element 1. The reception output of the ultrasonic element 1 is amplified by the preamplifier 7 and the AGC 8, and then converted into digital data sampled and digitized by the AD converter 9, and stored in the storage unit 10 (step S207). Hereinafter, it is assumed that the digital data stored in step S207 is the signal f (t) in FIG. Note that the digital data stored in step S207 is not limited to the signal f (t) in FIG.

  Subsequently, the Fourier transform unit 112 performs a Fourier transform on the reception output (digital data) of the ultrasonic element 1 stored in the storage unit 10 in step S207, that is, the signal f (t) to generate a signal F (s). (Step S208). 6A shows the real part of the signal F (s), and FIG. 6B shows the imaginary part of the signal F (s). The Fourier transform unit 112 outputs the signal F (s) to the calculation unit 113.

  Subsequently, when the time T4 elapses after the output of the operation signal, the measurement instruction unit 111 sends a switching signal for selecting the ultrasonic element 1 by the multiplexer 5 and selecting the ultrasonic element 2 by the multiplexer 5 and 6. (Step S209). The time T4 is a time obtained by adding the specified time T5 to the time T1. The specified time T5 is a longer time than a maximum time required from when a 4 MHz ultrasonic signal is transmitted to when a 4 MHz ultrasonic signal is received.

  Upon receiving the switching signal, the multiplexer 5 selects the ultrasonic element 1 as the transmission side (step S210). Further, when receiving the switching signal, the multiplexer 6 selects the ultrasonic element 2 as the receiving side (step S211).

  The measurement instruction unit 111 outputs an operation signal to the transmission signal generator 3 after a specific time T3 from the output of the switching signal in Step 209 (Step S212).

  When receiving the operation signal, the transmission signal generator 3 outputs the transmission signal for the output time T1 (step S213).

  The transmission signal is amplified by the power amplifier 4 so as to match the excitation voltage of the ultrasonic element 1 and the ultrasonic element 2, and then reaches the ultrasonic element 1 via the multiplexer 5. The ultrasonic element 1 transmits an ultrasonic signal of 4 MHz toward the ultrasonic element 2 through the fluid to be measured in the flow path 200 in accordance with the transmission signal subjected to power amplification (step S214).

  On the other hand, the reception output of the ultrasonic element 2 is output to the preamplifier 7 after the multiplexer 6 selects the ultrasonic element 2. The reception output of the ultrasonic element 2 is amplified by the preamplifier 7 and the AGC 8, and then converted into digital data sampled and digitized by the AD converter 9, and stored in the storage unit 10 (step S215). Hereinafter, it is assumed that the digital data stored in step S215 is the signal g (t) in FIG. Note that the digital data stored in step S215 is not limited to the signal g (t) in FIG.

  Subsequently, the Fourier transform unit 112 performs a Fourier transform on the reception output (digital data) of the ultrasonic element 2 stored in the storage unit 10 in step S215, that is, the signal g (t), and generates the signal G (s). (Step S216). 6A shows the real part of the signal G (s), and FIG. 6B shows the imaginary part of the signal G (s).

  The measurement instructing unit 111 stops the multiplexer 5 from selecting the ultrasonic element 1 and the multiplexer 6 from selecting the ultrasonic element 2 after the time T4 has elapsed since the operation signal was output in step S212. Are output to the multiplexers 5 and 6.

  The Fourier transform unit 112 outputs the signal G (s) to the calculation unit 113.

  When receiving the signal F (s) and the signal G (s), the arithmetic unit 113 calculates a signal H (s) that is a conjugate complex signal of the signal G (s) (step S217). Specifically, the calculation unit 113 calculates the signal H (s) by inverting the sign of the imaginary part (Imaginal part) of the signal G (s).

  Subsequently, the calculation unit 113 calculates the inner product of the signal F (s) and the signal H (s) (step S218). The calculation unit 113 outputs the calculation result of the inner product of the signal F (s) and the signal H (s) to the inverse Fourier transform unit 114.

  When receiving the calculation result of the inner product of the signal F (s) and the signal H (s), the inverse Fourier transform unit 114 performs an inverse Fourier transform on the calculation result to generate a signal c (t) (step S219). ). FIG. 7 shows the real part of the signal c (t). The inverse Fourier transform unit 114 outputs the signal c (t) to the calculation unit 115.

  As shown in FIG. 7, the calculation unit 115 specifies the timing ta at which the maximum value is given in the real part of the signal c (t) (step S220). The timing ta is a time tu until the ultrasonic signal transmitted from the upstream ultrasonic element 1 reaches the downstream ultrasonic element 2 and the ultrasonic wave transmitted from the downstream ultrasonic element 2. This represents a difference Δt (= td−tu) from the time td until the signal reaches the upstream ultrasonic element 1.

  Subsequently, the calculation unit 115 calculates the flow rate of the fluid to be measured based on the timing ta (step S221).

  Note that the flow rate of the fluid to be measured is obtained by multiplying the average flow velocity (m / sec) of the fluid to be measured by the cross-sectional area of the flow path 200. Since the cross-sectional area of the channel 200 is a value (constant) that can be measured in advance, if the average flow velocity (m / sec) of the fluid to be measured is obtained, the calculation unit 115 can calculate the flow rate of the fluid to be measured. it can.

  Here, the length of the ultrasonic propagation path in the fluid to be measured is L, the angle of the ultrasonic wave propagation direction with respect to the flow direction of the fluid to be measured is θ, and the velocity of the ultrasonic wave passing through the stationary fluid is c Then, the average flow velocity V of the fluid to be measured is expressed as follows.

  In Equation 1, since L, θ, and c are constants, the average flow velocity V can be obtained by obtaining Δt in the formula on the left side.

  In Equation 1, in the expression on the right side, instead of the constant c, the time tu until the ultrasonic signal transmitted from the upstream ultrasonic element 1 reaches the downstream ultrasonic element 2 and the downstream super The time td until the ultrasonic signal transmitted from the acoustic element 2 reaches the upstream ultrasonic element 1 is used. Since the measurement method of the time tu and the time td is known, a detailed description is omitted. For example, the calculation unit 115 stores the timing at which the measurement instruction unit 111 outputs the operation signal and the storage unit 10. It measures using the timing of the reception output of the ultrasonic elements 1 and 2.

  The calculation unit 115 calculates the average flow velocity V according to the equation (1) using the timing ta as Δt, and calculates the flow rate of the fluid to be measured by multiplying the average flow velocity V by the cross-sectional area (constant) of the flow path 200.

  The calculation unit 115 displays the calculated flow rate of the fluid to be measured on the display unit 12. Note that the calculation unit 115 may display the timing ta on the display unit 12.

  Next, the effect of this embodiment will be described.

  According to the present embodiment, Fourier transform is used to calculate a cross-correlation function (signal c (t)) between the signal f (t) and the signal g (t). In the Fourier transform, calculation using all of the signal f (t) and calculation using all of the signal g (t) are performed. For this reason, the cross-correlation function calculated using the Fourier transform is less susceptible to noise at specific measurement points of the signal f (t) and the signal g (t). Therefore, it is possible to suppress the noise at a specific measurement point from greatly affecting the measurement accuracy.

  In addition, in the conventional method (zero cross method or product-sum operation method), only about 1/10 time accuracy is obtained with respect to the sampling period, but in this embodiment, it is more than 1/100 of the sampling period. Even better measurement accuracy can be obtained.

  In the present embodiment, a predetermined ultrasonic signal is transmitted and received between the pair of ultrasonic elements 1 and 2, and Fourier transform is performed on the reception output of each ultrasonic element 1 and 2. In the Fourier transform, processing using a window function is often used. In this embodiment, the shape of the reception output of each ultrasonic element can be adjusted by the transmitted ultrasonic signal. Therefore, for example, as an ultrasonic signal to be transmitted, a signal in which the reception output itself of each ultrasonic element 1 is processed by a window function (for example, a waveform similar to the waveform of the signal f (t) in FIG. 5). If an ultrasonic signal having (2) is used, processing using a window function can be made unnecessary. Note that an ultrasonic signal having a waveform similar to the waveform of the signal f (t) in FIG. 5 is easily realized by starting the supply of the excitation voltage to the ultrasonic transducer and then stopping the supply of the excitation voltage. it can.

  In the case of the method of multiplying the window function before the DFT calculation processing, changing the start position and end position of the window function changes the value of the time difference that is the calculation result. Furthermore, it cannot be said strictly that the sampling data is weighted by multiplying by the window function and the time difference between the data points obtained by the sampling is calculated. If processing using a window function can be eliminated, these problems can be solved.

  In the present embodiment, an ultrasonic signal having a predetermined frequency (for example, 4 MHz) is used. For this reason, for example, when a filter for reception output is provided, a high-accuracy filter that passes a single-frequency ultrasonic signal is easy to use. Further, when Fourier transform (for example, DFT) is performed on an ultrasonic signal having a single frequency, a noise component is pushed to a frequency other than the main component frequency (frequency of the transmitted ultrasonic signal). When calculating the conjugate complex product (inner product), the value of the frequency component other than the main component frequency is negligibly small. As a result, the main component frequency in the two (upstream and downstream) receptions is calculated. The items involved are weighted and calculated.

  If the signal f (t) and the signal g (t) are directly subjected to the DFT calculation process, a large noise is generated when the signal values at the calculation start point and the end point are different as the characteristics of the Fourier transform. In addition, noise components included in the signal f (t) and the signal g (t) deteriorate the accuracy of time difference calculation.

  For this reason, for example, the Fourier transform unit 112 forces the part of the signal f (t) or the signal g (t) before or after the ultrasonic signal part (received signal of the ultrasonic signal) that is the target of the time difference comparison. Therefore, it is desirable to perform signal processing of “0” and perform DFT on each of the signal f (t) and signal g (t) on which the signal processing has been performed. This signal processing may be performed by a processing unit (not shown) provided in the previous stage of the Fourier transform unit 112.

  By performing this signal processing, it is possible to reduce the noise problem caused by the DFT arithmetic processing and improve the accuracy of time difference calculation. “0” is an example of a specific value. The specific value is not limited to “0” and can be changed as appropriate.

  In addition, calculation of Fourier series in DFT calculation and IDFT calculation takes a lot of time because it includes many terms. The values of the sin function and the cos function necessary for the calculation of the Fourier series are in the following form.

Where k, n is an integer, typically k, n = 0 to N−1
Therefore, in the case of k, n = 0 to N−1, the values of the sin function and the cos function are calculated in advance and stored as a table (storage means), and the Fourier transform unit 112 calculates DFT. When the inverse Fourier transform unit 114 calculates IDFT, the sin function value and the cos function value necessary for the Fourier transform and the inverse Fourier transform are reconstructed by referring to the table with k and n as arguments as shown in Equation 3. It may be possible to calculate DFT or IDFT at high speed without performing calculation.

  FIG. 8 is a diagram showing an example of an ultrasonic flowmeter provided with a storage unit 13 storing a table showing sin function values and cos function values necessary for Fourier transform and inverse Fourier transform. In FIG. 8, the same components as those shown in FIG.

  In the ultrasonic flowmeter shown in FIG. 8, the Fourier transform unit 112 performs Fourier transform with reference to a table (sin function value and cos function value necessary for Fourier transform) in the storage unit 13. Further, the inverse Fourier transform unit 114 performs the Fourier transform with reference to a table in the storage unit 13 (sin function values and cos function values necessary for the inverse Fourier transform).

  Moreover, in the ultrasonic flowmeter shown in FIGS. 1 and 8, the following method may be used as a method for obtaining the maximum value timing ta of the signal c (t).

(1) First method: When performing IDFT operation using a coefficient sequence obtained by DFT operation and conjugate complex product, the target effective digit is obtained by repeatedly calculating the IDFT operation value not only at the sampling point but also at any point. A method to obtain the time difference between numbers ”
In the conventional product-sum operation method, the point at which the product-sum value obtained by shifting the time in units of sampling points is maximized is obtained. For this reason, in the product-sum operation method, only time difference accuracy in units of sampling periods can be obtained.

  In the first method, the inverse Fourier transform unit 114 first performs an IDFT operation on the cross spectrum of the signal f (t) and the signal g (t) in units of sampling points T2 (first time interval), and performs a preliminary operation. A signal (seventh signal) is generated.

  Subsequently, the inverse Fourier transform unit 114 obtains the maximum value timing (the most likely time difference) tb of the spare signal based on the sampling period.

  Subsequently, the inverse Fourier transform unit 114, in a time range including before and after the timing tb (for example, a range from “tb−α” to “tb + α”), is a time interval shorter than the sampling point interval (a time difference interval with a required accuracy). ), The signal c (t) is generated by performing the IDFT operation on the cross spectrum of the signal f (t) and the signal g (t). The inverse Fourier transform unit 114 outputs this signal c (t) to the calculation unit 115. The calculation unit 115 calculates the flow rate of the fluid to be measured based on the maximum value timing in the signal c (t).

  For this reason, according to the first method, the most probable time difference Δt can be calculated with the required accuracy.

(2) Second method “A method for obtaining a point where the differential value of the function of the IDFT calculation result is 0 when calculating the point where the DFT calculation result is the maximum, that is, the target time difference” by linear interpolation.
In order to find the point where the value of the real part of the function of the IDFT calculation result is maximized, it takes time to perform recalculation by increasing / decreasing arguments until the target precision is reached.

  In the second method, a formula that gives the derivative value of the function of the IDFT operation result is derived, and the fact that the derivative value of the function of the IDFT operation result is 0 in terms of giving the maximum value of the function of the IDFT operation result. By calculating the point that gives the maximum value of the function of the IDFT calculation result, the timing of the maximum value can be calculated without repeating the calculation.

  For example, the inverse Fourier transform unit 114 calculates the signal c (t) by performing IDFT calculation on the cross spectrum of the signal f (t) and the signal g (t) in advance at each sampling point (predetermined time interval). To do.

  Subsequently, the calculation unit 115 specifies the timing of the maximum value (local maximum value) of the signal c (t) (sampling point that gives the largest IDFT value).

  Subsequently, the calculation unit 115 has a differential value of the function of the IDFT calculation result on the sampling point immediately before the specified timing as a slope, and a straight line passing through the IDFT calculation result, and immediately after the specified timing. Using the straight line passing through the IDFT calculation result with the derivative value of the function of the IDFT calculation result on the sampling point as the slope, the signal c (t) between the previous sampling point and the next sampling point is a straight line Interpolate.

  Subsequently, the calculation unit 115 specifies the timing at which the maximum value of the linearly interpolated signal c (t) is taken (timing corresponding to the intersection of the straight lines) as the timing at which the differential value of the signal c (t) becomes zero. .

  It should be noted that the target accuracy can be obtained if the interval between the sampling points is sufficiently narrow so that the target accuracy is obtained.

  The calculation formula for the real part of IDFT at arbitrary time n has the following form.

  In Equation 4, CCR_real represents the real part of the cross spectrum of the signal f (t) and the signal g (t), and CCR_imag represents the imaginary part of the cross spectrum of the signal f (t) and the signal g (t). Show.

  Differentiating IDFT with respect to n yields:

  here,

If n0 to be obtained is between n-1 and n + 1 and linear interpolation can be performed, n0 to be obtained can be calculated as follows.

  The calculation unit 115 calculates n0 (timing for taking the maximum value of the linearly interpolated signal c (t)) according to Equation 7.

  In the above-described embodiment, the signal f (t) is an upstream signal in a period from when the downstream ultrasonic element 2 transmits an ultrasonic signal until the upstream ultrasonic element 1 finishes receiving the ultrasonic signal. The signal g (t) may be output from the upstream ultrasonic element 1 after transmitting the ultrasonic signal until the downstream ultrasonic element 2 finishes receiving the ultrasonic signal. It may be the output of the ultrasonic element 2 on the downstream side during this period.

  In the above-described embodiment, each unit in the processing unit 11 may be realized by the processing unit 11 reading and executing a computer program recorded on a computer-readable recording medium, or realized by individual hardware. May be.

  In the embodiment described above, the illustrated configuration is merely an example, and the present invention is not limited to the configuration.

1, 2 Ultrasonic element 3 Transmission signal generator 4 Power amplifier 5, 6 Multiplexer 7 Preamplifier 8 AGC
9 AD converter 10 RAM
DESCRIPTION OF SYMBOLS 11 Processing part 111 Measurement instruction | indication part 112 Fourier-transform part 113 Calculation part 114 Inverse Fourier-transform part 115 Calculation part 200 Flow path

Claims (5)

A pair of ultrasonic elements that are arranged opposite to each other across the flow path on the upstream side and downstream side of the flow path through which the fluid to be measured flows, and transmit and receive a predetermined ultrasonic signal;
For each of the first signal that is the reception output of the upstream ultrasonic element and the second signal that is the reception output of the downstream ultrasonic element, the third signal and the fourth signal are subjected to Fourier transform. Fourier transform means for generating
Arithmetic means for generating a conjugate complex signal of one of the third signal and the fourth signal, calculating an inner product of the other signal and the conjugate complex signal, and generating a fifth signal;
An inverse Fourier transform means for performing an inverse Fourier transform on the fifth signal to generate a sixth signal;
Based on the timing of the maximum value in said sixth signal, seen including a calculation unit, a for calculating the flow rate of the fluid to be measured,
The inverse Fourier transform means generates the sixth signal by performing the inverse Fourier transform on the fifth signal at a predetermined time interval,
The calculating means specifies the timing of the maximum value in the sixth signal, and includes a straight line having a derivative value of the sixth signal at a timing before the predetermined time as a slope from the timing of the maximum value, and the maximum value A straight line having a slope of a differential value of the sixth signal at the timing after the predetermined time from the timing, and the sixth signal between the timing before the predetermined time and the timing after the predetermined time is a straight line. An ultrasonic flowmeter that interpolates and calculates the timing of the maximum value based on a result of the linear interpolation . The ultrasonic flowmeter according to claim 1,
The ultrasonic flowmeter, wherein the predetermined ultrasonic signal is an ultrasonic signal having a predetermined frequency. The ultrasonic flowmeter according to claim 1 or 2,
The Fourier transform means performs signal processing with specific values at both ends of the first signal and both ends of the second signal, and for each of the first signal and the second signal subjected to the signal processing, An ultrasonic flowmeter that performs Fourier transform to generate the third signal and the fourth signal. The ultrasonic flowmeter according to any one of claims 1 to 3,
Storage means for storing a table showing sin function values and cos function values necessary for the Fourier transform and the inverse Fourier transform;
The Fourier transform means performs the Fourier transform with reference to the table,
The inverse Fourier transform unit is an ultrasonic flowmeter that performs the inverse Fourier transform with reference to the table. In the ultrasonic flowmeter according to any one of claims 1 to 4,
The inverse Fourier transform means generates the seventh signal by performing the inverse Fourier transform on the fifth signal at a first time interval, and for the predetermined time region including before and after the maximum value timing of the seventh signal. An ultrasonic flowmeter that generates the sixth signal by performing the inverse Fourier transform on five signals at a second time interval shorter than the first time interval.

Images