TWI732351B - Tube flow rate measurement system and method thereof - Google Patents

Abstract

A tube flow rate measurement system is configured to measure a fluid flow rate of a fluid passing through a test segment of a tube, and the fluid flow rate measurement system comprises a first sensor set, a second sensor set, and a computing device. The first sensor set and the second sensor set are respectively configured to connect with two ends of the test segment, and the computing device is electrically connected to the first sensor set and the second sensor set. The computing device calculates a fluid velocity set according to a plurality of sensing signal pairs transmitted between the first sensor set and the second sensor set, and the computing device stores a plurality of classifications which is generated by executing a supervised learning many times and a fluid flow rate correction coefficient database. The computing device determines a flow field type of the fluid velocity set according to the classifications, and obtains a fluid flow rate correction coefficient corresponding to the flow field type of the fluid velocity set from the fluid flow rate correction coefficient database, and calculate the fluid flow rate according to the fluid flow rate correction coefficient and the fluid velocity set.

Description

Pipeline flow measurement system and method

The invention relates to a measuring system and a measuring method for the flow of fluid in a pipeline.

The current central air-conditioning system involves the regulation of heat transfer, water volume, air volume, and refrigerant volume between multiple devices. If the operation of multiple devices is poorly coordinated, electrical energy will be wasted. Among the various equipment of the central air-conditioning system, the water-side system consumes the most power. Therefore, the efficiency of the water-side system has a decisive influence on the energy-saving efficiency of the central air-conditioning system. The water-side system includes an ice water main engine, an ice water pump, and a cooling water tower. The operating performance of the water-side system is usually judged based on kW/RT, where kW is the total power consumption of the water-side system, and RT is based on the flow of ice water and Supply refrigerated tons calculated from the temperature difference between the inlet and outlet of ice water. When the value of kW/RT is smaller, it means that the efficiency of the water side system is better, which in turn affects the energy-saving efficiency of the central air conditioning system.

As the piping design of the water-side system is becoming more and more complex, the fluid in many short-length pipes cannot form a fully developed flow. However, the current flow measurement technology is not suitable for non-fully developed flow. The error of the fully developed flow measurement can even reach 10%, which makes it impossible to accurately measure the performance of the central air-conditioning water-side system, which makes it difficult to implement energy-saving improvement measures and difficult to verify the effectiveness. At present, domestic commercial and office buildings usually use central air-conditioning systems, making the power consumption of central air-conditioning systems the largest domestic electricity consumption. In response to the current problem of frequent overloading of power generation in power plants, it is necessary to accurately control the flow and flow of water-side systems. Efficiency to achieve the effect of energy saving and carbon reduction.

The present invention provides a pipeline flow measurement system and method. Regardless of whether the fluid in the pipeline belongs to an incompletely developed flow or a fully developed flow, the flow of the fluid in the pipeline can be accurately measured, thereby achieving energy saving and carbon reduction Effect.

A pipeline flow measurement system disclosed in the present invention is used to measure the flow of fluid in a section to be measured of a pipeline. The pipeline flow measurement system includes a first sensor group and a second sensor The first sensor group and the second sensor group are respectively used to connect to two different ends of the segment to be measured, and the computing device is electrically connected to the first sensor group and the second sensor group. Detector group. The arithmetic device calculates a flow rate combination based on the multiple sensor signal pairs transmitted between the first sensor group and the second sensor group, and stores multiple classification algorithms generated by performing multiple supervised learning And flow correction coefficient database. The computing device determines the flow field type to which the flow velocity combination belongs according to the classification algorithms, obtains the flow correction coefficient corresponding to the flow field type from the flow correction coefficient database, and calculates the flow rate of the fluid based on the flow correction coefficient and the flow velocity combination .

The present invention discloses a method for measuring the flow rate of a pipeline, which is used to measure the flow rate of a fluid in a section to be measured of a pipeline, including: reading and transmitting to a first sensor group and a second sensor by an arithmetic device The multiple sensing signal pairs between the sensor groups; the calculation device calculates a flow rate combination of the fluid according to the sense signal pairs, wherein the calculation device stores multiple classification calculations generated by performing multiple supervised learning Method and flow correction coefficient database; use these classification algorithms to determine the flow field type to which the flow rate combination belongs; use the flow field type to obtain the flow correction coefficient corresponding to the flow field type from the flow correction coefficient database; and use the calculation device to base it The flow rate is combined with the flow correction coefficient to calculate the flow rate of the fluid.

The pipeline flow measurement system and the method for measuring pipeline flow disclosed in the present invention, for example, through multiple executions of multiple different classification algorithms generated by a support vector machine belonging to supervised learning, even if the fluid in the pipeline If it belongs to an incompletely developed flow, the flow field type of the fluid can also be accurately determined according to the hyperplane equations, so as to obtain an appropriate flow correction coefficient from the correction coefficient database according to the flow field type. In this way, the arithmetic device can be based on the flow of the fluid Speed data to accurately determine the type of flow field the fluid belongs to, find out the corresponding flow correction coefficient based on the flow field type, calculate more accurate fluid flow based on the flow correction coefficient, and effectively monitor the water side based on the accurate flow value The efficiency of the system, in turn, achieves the effect of energy saving and carbon reduction.

The above description of the disclosure and the following description of the embodiments are used to demonstrate and explain the spirit and principle of the present invention, and to provide a further explanation of the scope of the patent application of the present invention.

11: The first sensor group

111~114: The first sensor

12: The second sensor group

121~124: second sensor

13: Computing device

131: Processor

132: Memory

1321: Classification database of flow field types

1322: Flow correction coefficient database

133: Controller

134: Input operation interface

T: Pipeline

T1: segment to be tested

Ex: Entry side

En: export side

P1~P8: the first position to the eighth position

L: central axis

N: cross section

C1~C16: Measurement path

HP1: the first hyperplane

HP2: second hyperplane

V1: First flow rate

V2: Second flow rate

f1: the first flow field

f2: second flow field

f3: third flow field

FIG. 1 is a functional block diagram of the first embodiment of the pipeline flow measurement system disclosed in the present invention.

2 is a schematic diagram showing a first embodiment of the pipeline flow measurement system disclosed in the present invention applied to pipelines.

FIG. 3 is a schematic diagram of FIG. 2 from another perspective.

4 is a schematic diagram showing a second embodiment of the pipeline flow measurement system disclosed in the present invention applied to pipelines.

FIG. 5 is a schematic diagram showing a third embodiment of the pipeline flow measurement system disclosed in the present invention applied to pipelines.

FIG. 6 is a method flowchart of an embodiment of the method for measuring pipeline flow rate disclosed in the present invention.

FIG. 7 is a detailed flowchart of determining the type of flow field of the fluid in the pipeline disclosed in FIG. 6.

FIG. 8 is a schematic diagram illustrating the type of flow field to which the fluid in the pipeline belongs to the fluid in the pipeline disclosed in FIG. 6.

FIG. 9 is a flowchart of a method for establishing a flow correction coefficient according to an embodiment of the present invention.

FIG. 10 is a flowchart of an embodiment of the present invention executing a machine learning program to build a flow field classification database.

The detailed features and advantages of the present invention will be described in detail in the following embodiments. The content is sufficient to enable anyone familiar with the relevant art to understand the technical content of the present invention and implement it accordingly, and according to the content disclosed in this specification, the scope of patent application and the drawings. Anyone who is familiar with relevant skills can easily understand the purpose and advantages of the present invention. The following examples further illustrate the viewpoints of the present invention in detail, but do not limit the scope of the present invention by any viewpoint.

FIG. 1 is a functional block diagram of the first embodiment of the pipeline flow measurement system disclosed in the present invention. As shown in FIG. 1, the pipeline flow measurement system includes a first sensor group 11, a second sensor group 12, and an arithmetic device 13, and the arithmetic device 13 is electrically connected to the first sensor group 11 and the second sensor group 12. The computing device 13 is, for example, a cloud host, a server, or a mobile communication device. The computing device includes a processor 131, a memory 132, a controller 133, and an input operation interface 134, and the processor 131 is electrically connected to the memory 132 , A controller 133 and an input operation interface 134. The memory 132 stores a plurality of first-class field classification databases 1321 and a flow correction coefficient database 1322. The flow field classification database 1321 includes multiple classification calculations generated by performing supervised learning multiple times. The classification algorithm is used to classify a plurality of different flow field types, and the flow correction coefficient database 1321 contains a plurality of different flow field types and a plurality of different flow field types corresponding to the flow field types. Flow correction factor. In this embodiment, the supervised learning is an example of a support vector machine. Each time the support vector machine is executed, a hyperplane equation can be generated to classify two flow field types. By executing the support vector machine multiple times A plurality of different hyperplane equations (hyperplane equations) can be generated to classify a plurality of two kinds of flow field types. For example, the flow field classification database 1321 includes a first hyperplane equation and a second hyperplane equation generated by the execution of a support vector machine. The first hyperplane equation is used to determine whether the fluid model belongs to the flow field A, and the second hyperplane equation is used to determine whether the fluid model belongs to the flow field A. The hyperplane equation is used to determine whether the flow field model belongs to the B flow field. The flow correction coefficient database 1322 includes flow field A and its corresponding flow correction coefficient D1 (for example, 1.0), and flow field B and its corresponding flow correction coefficient D2 (for example, 0.955).

The user inputs a start command through the input operation interface 134, and then the input operation interface 134 will send the start command to the processor 131, and the processor 131 sends the start command to the controller 133, and the controller 133 enables the first command according to the start command. A sensor group 11 and a second sensor group 12, in this way, the first sensor group 11 and the second sensor group 12 can sense the state of the fluid in the pipeline to generate a sensing signal, The sensing signal is sent to the controller 133, and the controller 133 transmits the sensing signal to the processor 131 for subsequent flow calculation.

FIG. 2 is a schematic diagram of a first embodiment of the pipeline flow measurement system disclosed in the present invention applied to a pipeline, and FIG. 3 is a schematic diagram of FIG. 2 from another perspective. As shown in FIG. 2, the pipeline flow measurement system is used to measure the flow of fluid in the section to be measured of the pipeline T, where the section to be measured can be the pipeline T or a part of the pipeline T. In this embodiment, the section to be measured is the pipeline T, so the different ends of the section to be measured are the inlet end Ex and the outlet end En of the pipeline T, and the first sensor group 11 and the second sensor group 11 The sensor group 12 is connected to the inlet end Ex and the outlet end En of the pipeline T, respectively.

The flow velocity of a fully developed fluid is evenly distributed, and only one flow velocity data is needed to accurately represent the actual state of the fluid. However, the flow velocity distribution of the incompletely developed flow is not uniform. One flow velocity data cannot accurately represent the actual state of the fluid, and multiple flow velocity data must be used to correctly represent the actual state of the fluid. In order to accurately represent the actual state of the incompletely developed fluid, a sufficient number of sensors must be installed on the pipeline. In this embodiment, the first sensor group 11 includes a plurality of first sensors 111 to 114, and the second sensor group 12 includes a plurality of second sensors 121 to 124. In this embodiment, the first sensors 111 to 114 and the second sensors 121 to 124 are ultrasonic sensors. The inlet end Ex of the pipeline T is provided with a first position to a fourth position P1~P4, and the distance between the first position to the fourth position P1~P4 is equal to each other and is located outside the pipeline T, and these The first sensors 111-114 are respectively assembled in the first position to the fourth position P1 to P4. The outlet end En of the pipeline T is provided with the fifth position to the eighth position P5~P8, and the distance between the fifth position to the eighth position P5~P8 is equal, and the fifth position to the eighth position P5~P8 Located outside the pipeline T, and the second sensors 121~124 are respectively assembled to The fifth position to the eighth position P5~P8. The pipeline T defines a central axis L and a section N passing through the central axis L, and the central axis L is located on the section N. The first sensors 111 to 114 are located on the left side of the section N, and the second sensors 121 to 124 are located on the right side of the section N. The memory 132 further stores a weighted combination, wherein the weighted combination is related to the arrangement positions of the first sensor group 11 and the second sensor group 12 in the pipeline T and the diameter of the pipeline T.

When the first sensor group 11 is used as a signal sending end, only one first sensor sends a sensing signal at the same time, which can be received by the second sensors. When the second sensor group 12 is used as a signal transmitter, only one second sensor sends a sensing signal at the same time, which can be received by the first sensors. As shown in FIG. 3, there are different measurement paths C1 to C4 between the first sensor 111 and the second sensors 121 to 124, and the first sensor 112 and the second sensors 121 There are different measurement paths C5~C8 between ~124, the first sensor 113 and the second sensors 121~124 have different measurement paths C9~C12, and the first sensor 114 There are different measurement paths C13 to C16 between the second sensors 121 to 124. For example, when the first sensor 111 sends the first sensing signal, the other first sensors 112 to 114 will not send any sensing signal, and the second sensors 121 to 124 respectively pass through The measurement paths c1~c4 receive the first sensing signal. When the second sensor 121 transmits the second sensing signal, the other second sensors 122 to 124 do not transmit the second sensing signal, and the first sensors 111 to 114 respectively pass the measurement path C1, C5, C9 and C13 receive the second sensing signal. A first sensing signal and a second sensing signal transmitted on the same measurement path (such as C1) are defined as a sensing signal pair. In this embodiment, the first sensor and the second sensor Each has four, so 16 sensing signal pairs can be formed. In this way, the processor 131 calculates a flow rate combination based on 16 pairs of sensing signals transmitted between the first sensor group 11 and the second sensor group 12, and the flow rate combination includes 16 flow rate data. The formula used by the processor 131 to calculate the flow velocity v is derived as follows: c-vcos θ =L/t1 (Equation 1).

c+vcos θ =L/t2 (Equation 2).

2vcos θ = L(1/t2-1/t1) (Equation 3).

v=(t1-t2)/2cos θ *(L/(t1*t2)) (Equation 4).

In the above equation, the meaning of each parameter is listed below. c: sound velocity, L: measurement path length, θ: sound wave transmission angle, v: flow velocity, t1: sound wave transmission time under countercurrent state, t2: sound wave transmission time under downstream state.

The processor 131 determines the flow field type to which the measured flow velocity combination belongs according to the hyperplane equations. Then the processor 131 obtains the corresponding flow correction coefficient from the flow correction coefficient database 1321 according to the confirmed flow field type. The processor 131 calculates the flow rate of the fluid according to the flow correction coefficient, the flow rate combination and the weight combination, and the processor 131 calculates the flow rate of the fluid with the following formula: Q=D*Vset*Wset (Equation 5).

In the above formula 5, the meaning of each parameter is listed below. Q: flow, D: flow correction coefficient, Vset: flow rate combination, Wset: weight combination.

4 is a schematic diagram showing a second embodiment of the pipeline flow measurement system disclosed in the present invention applied to pipelines. The difference between Fig. 4 and Fig. 2 is that the test section T1 is a part of the pipeline T, the test section T1 is between the inlet end Ex and the outlet end En of the pipeline T, and the first sensor group 11 and The second sensor group 12 is respectively grouped and connected to two different ends of the section T1 to be measured. As a result, the measurement path between the first sensor group 11 and the second sensor group 12 will be different from the measurement path of the first embodiment.

FIG. 5 is a schematic diagram showing a third embodiment of the pipeline flow measurement system disclosed in the present invention applied to pipelines. The difference between FIG. 5 and FIG. 2 is that the number of the first sensor provided at the inlet end Ex of the pipeline T is one. In this way, there are four pairs of sensing signals transmitted between the inlet end Ex of the pipeline T and the outlet end En of the pipeline T, and the corresponding flow velocity combination includes four flow velocity data.

In other embodiments, the number of the second sensor provided at the outlet end En of the pipeline T is one, while the number of the first sensor provided at the inlet end Ex of the pipeline T remains large. A. In addition, the pipeline T is not limited to a straight pipe, and may be a curved pipe.

FIG. 6 is a method flowchart of the first embodiment of the method for measuring pipeline flow rate disclosed in the present invention, and any embodiment of the aforementioned pipeline flow rate measurement system can execute the method of FIG. 6. As shown in FIG. 6, in step S101, the controller 133 respectively enables the first sensor group 11 and the second sensor group 12 connected to different ends of the section to be measured of the pipeline T. In step S102, the processor 131 reads a plurality of sensing signal pairs transmitted between the first sensor group 11 and the second sensor group 12, wherein the number of the sensing signal pairs depends on the first sensor The number of sensors and the number of second sensors. In step S103, the processor 131 calculates a flow velocity combination in the pipeline to be measured according to the sensing signal pairs, and the flow velocity combination includes a plurality of flow velocity data, wherein the number of the flow velocity data is equal to the number of the sensing signals. The number of signal pairs. In step S104, the processor 131 determines the flow field type to which the combination of the flow velocity of the fluid in the test section of the pipeline belongs according to a plurality of different hyperplane equations generated by multiple executions of the support vector machine. In step S105, the processor 131 obtains the flow correction coefficient corresponding to the flow field type from the flow correction coefficient database 1321 according to the confirmed flow field type of the fluid. In step S106, the processor 131 calculates the product of the flow rate combination, the weight combination, and the flow correction coefficient to obtain the flow rate of the fluid in the section to be measured of the pipeline T.

In other embodiments, a supervised learning different from a support vector machine can be used to generate multiple classification algorithms, and the classification algorithms can be used to determine the flow rate combination of the fluid in the pipeline to be measured. Field type.

FIG. 7 is a detailed flowchart for determining the type of flow field of the fluid in the pipeline disclosed in FIG. 6, and FIG. 8 is a diagram showing the type of flow field of determining the fluid in the pipeline disclosed in FIG. 6 Schematic diagram. As shown in Figure 7, in step S1041, for example, according to the polynomial kernel function of the support vector machine, the measured flow velocity is combined to project the kernel space, and the core space includes The multiple regional spaces distinguished by the hyperplane equations correspond to different flow field types, and the hyperplanes represented by the hyperplane equations do not intersect each other. In step S1042, the processor 131 determines the positional relationship between the flow velocity combination and the hyperplane equations. In steps S1043: Use the processor 131 to determine the regional space where the flow rate combination is located.

As shown in FIG. 8, the physical quantities of the ordinate and the abscissa can be any two flow rates (such as the first flow velocity V1 and the second flow velocity V2) in the flow velocity combination (such as (V1, V2...V16)). The core space contains a first hyperplane HP1 and a second hyperplane HP2. In the first area enclosed by the first hyperplane HP1, multiple fluid models belonging to the second flow field f2 are distributed. The second hyperplane HP1 Multiple fluid models belonging to the third flow field f3 are distributed in the second area space enclosed by the plane HP2, and multiple fluid models belonging to the first flow field f1 are distributed outside the first area space and outside the second area space. A fluid model. When the processor 131 determines that the flow velocity combination in the pipeline T is located in the first regional space, the flow field type of the fluid belongs to the second flow field f2.

FIG. 9 is a flowchart of a method for establishing a flow correction coefficient according to an embodiment of the present invention. As shown in FIG. 9, in step S201: Measure an actual flow rate of a fluid model in the section to be measured in the pipeline through a flow meter. In step S202: use the first sensor group 11 and the second sensor group 12 to measure multiple pairs of sensing signals in the segment to be measured. In step S203: the processor 131 calculates a flow rate combination in the section to be measured according to the sensing signal pairs, the flow rate combination includes a plurality of first type flow rates and a plurality of second type flow rates, wherein the first type flow rate The transmission direction and the cross section passing through the central axis of the pipeline are perpendicular to each other (meaning horizontal flow velocity), and the weight combination includes multiple first-type weights and multiple second-type weights, and the first-type weights correspond to the first-type weights. Type flow rate, the second type weights correspond to the second type flow rates. In step S204, the processor 131 calculates the product of the first-type flow rates and the first-type weights to obtain an estimated flow rate. In step S205, the processor 131 divides the estimated flow rate by the actual flow rate to obtain a flow rate correction coefficient. According to the method of establishing flow correction coefficient in Fig. 9, calculate the actual flow and estimated flow field for other different fluid models, and then the flow correction coefficient database can be established.

Due to the uneven distribution of the flow velocity of the incompletely developed flow, it is difficult to use manual classification or general linear regression to accurately classify the incompletely developed flow. Therefore, before the arithmetic device 3 reads the sensing signal pairs, a machine learning program is executed to build the flow field classification database 1321, so as to achieve the purpose of classifying the fluids of the incompletely developed flow.

FIG. 10 is a flowchart of an embodiment of executing a machine learning program to build a flow field classification database. As shown in Fig. 10, machine learning takes a support vector machine as an example. In step S301, a plurality of fluid model combinations of different flow field types are collected as a training sample, wherein each fluid model combination includes a plurality of different fluid models. Each fluid model is represented by multiple flow velocity data. For example, 8 different fluid models belonging to the A flow field are collected to form the first fluid model combination, and 10 different fluid models belonging to the B flow field are collected to form the second fluid model combination. In step S302, the homogeneity of the fluid models belonging to the same flow field type in the training sample is analyzed through the support vector machine. In step S303, a plurality of different hyper-plane equations are established according to the homogeneity analysis result, and the hyper-plane equations form a plurality of regional spaces, and the regional spaces correspond to different flow field types.

The parameter D in Table 1 is the diameter of the pipeline T, and the total length of the pipeline T is 10 times the diameter. The flow field type and flow rate are judged at the 2D, 4D, 6D, 8D, and 10D sections of the pipeline T. According to the calculation, the fluid circulating in the position of 2D~8D belongs to the incompletely developed flow. It can be seen from Table 1 that the pipeline flow measurement system and the method for measuring pipeline flow disclosed in the present invention effectively reduce the flow error to less than 2%, even if the total length of the pipeline T is only 2% of the diameter. The flow rate calculated according to the method for measuring pipeline flow rate provided by the present invention can meet actual requirements.

In summary, the pipeline flow measurement system and the method for measuring pipeline flow disclosed in the present invention have, for example, a classification database of flow field types established by multiple executions of support vector machines, which solves the problem of using manual classification in the past. Or the general linear regression method cannot classify the fluid of the incompletely developed flow. In this way, the computing device can accurately determine the type of flow field it belongs to based on the flow rate data, find the corresponding flow correction coefficient based on the flow field type, and calculate a more accurate fluid flow based on the flow correction coefficient, based on the accurate flow value The efficiency of the water side system can be monitored efficiently, and the effect of energy saving and carbon reduction can be achieved.

Although the present invention is disclosed in the foregoing embodiments, it is not intended to limit the present invention. All changes and modifications made without departing from the spirit and scope of the present invention fall within the scope of the patent protection of the present invention. For the scope of protection defined by the present invention, please refer to the attached scope of patent application.

11 first sensor group 12 second sensor group 13 arithmetic device 131 processor 132 memory 1321 Flow Field Classification Database 1322 flow correction coefficient database 133 controller 134 input operation interface

Claims (13)

A pipeline flow measurement system for measuring the flow of fluid in a section to be measured of a pipeline. The pipeline flow measurement system includes: a first sensor group and a second sensor group , The first sensor group and the second sensor group are respectively connected to two different ends of the segment to be measured; and an arithmetic device is electrically connected to the first sensor group and the second sensor group A sensor group, and calculate a flow rate combination according to a plurality of sensor signal pairs transmitted between the first sensor group and the second sensor group; wherein the computing device stores a combination of multiple executions A plurality of classification algorithms and a flow correction coefficient database generated by a supervised learning. The calculation device determines the flow field type to which the flow velocity combination belongs based on the classification algorithms. The calculation device uses the flow correction coefficient data The library obtains a flow correction coefficient corresponding to the type of flow field, and the computing device calculates the flow according to the flow correction coefficient and the flow velocity combination. The pipeline flow measurement system according to claim 1, wherein the supervised learning is a support vector machine, and the classification algorithms are a plurality of different hyperplane equations. The pipeline flow measurement system according to claim 1, wherein the calculation device obtains the flow rate by the product of the flow rate correction coefficient, the flow rate combination, and a weight combination. The pipeline flow measurement system according to claim 1, wherein the first sensor group includes a plurality of first sensors, the second sensor group includes a plurality of second sensors, and the to-be-tested The different ends of the segment are respectively an inlet end and an outlet end of the pipeline. The pipeline flow measurement system according to claim 4, wherein the pipeline defines a central axis and a cross section, and the central axis is located on the cross section, the first sensors are located on one side of the cross section, and The second sensors are located on the other side of the cross section. The pipeline flow measurement system according to claim 1, wherein the first sensor group and the second sensor group are assembled outside the pipeline, and the computing device is communicatively connected to the first sensor Sensor group and the second sensor group. The pipeline flow measurement system according to claim 1, wherein the computing device includes a processor, a controller, and a memory, the processor is electrically connected to the memory and the controller, and the memory stores the The classification algorithms and the flow correction coefficient database, the controller is electrically connected to the first sensor group and the second sensor group, and the controller is used to receive a start command to enable the first sensor Sensor group and the second sensor group. A method for measuring pipeline flow rate, used to measure the flow rate of fluid in a section to be measured of a pipeline, including: reading and transmitting to a first sensor group and a second sensor with an arithmetic device A plurality of sensing signal pairs between the device group; the calculation device calculates a flow velocity combination of the fluid according to the sensing signal pairs, wherein the calculation device stores a plurality of generated by performing a supervised learning multiple times Classification algorithm and a flow correction coefficient database; use these classification algorithms to determine the flow field type to which the flow velocity combination belongs; use the flow field type to obtain a flow correction corresponding to the flow field type from the flow correction coefficient database Coefficient; and the calculation device calculates the flow rate according to the flow rate combination and the flow rate correction coefficient. The method for measuring pipeline flow according to claim 8, wherein the supervised learning is a support vector machine, and the classification algorithms are a plurality of different hyperplane equations. The method for measuring pipeline flow according to claim 8, wherein the computing device obtains the flow by the product of the flow velocity combination, the flow correction coefficient, and a weight combination. The method for measuring pipeline flow as described in claim 8, further comprising executing a machine learning program before reading the sensing signal pairs by the computing device, the machine learning program including: collecting data of different flow field types A plurality of fluid module combinations are used as a training sample, and each fluid model combination includes a plurality of different fluid models; and the supervised learning is performed multiple times on the training sample to generate the classification algorithms. The method for measuring pipeline flow according to claim 8, wherein the first sensor group includes a plurality of first sensors, and the second sensor group includes a plurality of second sensors, each The first sensor emits a first sensing signal that is received by the second sensors, and each second sensor emits a second sensing signal that is received by the first sensors. The method for measuring pipeline flow according to claim 12, wherein the pipeline defines a central axis and a section passing through the central axis, and the central axis is located on the section, the first sensors and the A plurality of different measurement paths are included among the second sensors, and a part of the measurement paths passes through the central axis.

Images