CN117664227A - Multi-sensor integrated gas-liquid two-phase flow content measuring method - Google Patents
Multi-sensor integrated gas-liquid two-phase flow content measuring method Download PDFInfo
- Publication number
- CN117664227A CN117664227A CN202311628639.XA CN202311628639A CN117664227A CN 117664227 A CN117664227 A CN 117664227A CN 202311628639 A CN202311628639 A CN 202311628639A CN 117664227 A CN117664227 A CN 117664227A
- Authority
- CN
- China
- Prior art keywords
- gas
- liquid
- flow
- phase flow
- double
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 239000007788 liquid Substances 0.000 title claims abstract description 52
- 230000005514 two-phase flow Effects 0.000 title claims abstract description 51
- 238000000034 method Methods 0.000 title claims abstract description 20
- 230000004927 fusion Effects 0.000 claims abstract description 33
- 239000012530 fluid Substances 0.000 claims abstract description 15
- 238000000605 extraction Methods 0.000 claims abstract description 9
- 230000005284 excitation Effects 0.000 claims description 27
- 230000004913 activation Effects 0.000 claims description 15
- 238000012549 training Methods 0.000 claims description 15
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 13
- 230000005684 electric field Effects 0.000 claims description 9
- 238000011176 pooling Methods 0.000 claims description 9
- 238000012360 testing method Methods 0.000 claims description 9
- 238000005259 measurement Methods 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 6
- 238000000691 measurement method Methods 0.000 claims description 4
- 238000013528 artificial neural network Methods 0.000 claims description 3
- 230000000694 effects Effects 0.000 claims description 3
- 230000002708 enhancing effect Effects 0.000 claims description 3
- 238000002474 experimental method Methods 0.000 claims description 3
- 230000006698 induction Effects 0.000 claims description 3
- 238000003062 neural network model Methods 0.000 claims description 3
- 210000002569 neuron Anatomy 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 230000001360 synchronised effect Effects 0.000 claims description 3
- 230000006870 function Effects 0.000 description 14
- 230000009977 dual effect Effects 0.000 description 5
- 238000013136 deep learning model Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E30/00—Energy generation of nuclear origin
- Y02E30/30—Nuclear fission reactors
Landscapes
- Measuring Volume Flow (AREA)
Abstract
The invention relates to a gas-liquid two-phase flow content measuring method integrating multiple sensors. The integrated multiple sensors comprise a double-helix high-frequency microwave sensor and a precession vortex metering device, wherein the double-helix high-frequency microwave sensor is used for collecting signals of gas-liquid two-phase flow under different working conditions, fluid fluctuation information is measured based on the double-helix high-frequency microwave sensor, fluid flow and density information is measured based on the precession vortex metering device, and the double-flow fusion network is used for carrying out multi-sensor signal fusion and feature extraction and is used for measuring gas-liquid two-phase flow gas-content under different working conditions.
Description
Technical Field
The invention relates to a gas-liquid two-phase flow content measuring method, in particular to a gas-liquid two-phase flow content measuring method integrating multiple sensors.
Background
The gas-liquid two-phase flow has wide application in industrial fields, relates to a plurality of industries including petrochemical industry, energy sources, pharmacy, food processing and the like, and has application scenes including a gas-liquid separator, an oil-water separator, a pipeline conveying system and the like. In these applications, accurate measurement of gas-liquid two-phase flow parameters is critical to the stability, safety and economy of industrial production.
In recent years, the deep learning model is widely applied to sensor information fusion and parameter prediction because the deep learning model can extract elusive modes and information from complex sensor data, and can identify and utilize multi-level data characteristics, so that more accurate parameter measurement is realized compared with a traditional method. The gas-liquid two-phase flow complex parameter measurement method based on deep learning has the multiple advantages of high precision, real-time performance, data-driven decision, cost reduction, adaptability, automation and the like, and has wide application prospect.
Disclosure of Invention
The invention aims to provide a gas-liquid two-phase flow content measuring method integrating multiple sensors. The integrated multiple sensors comprise a double-helix high-frequency microwave sensor and a precession vortex metering device, wherein the double-helix high-frequency microwave sensor and the precession vortex metering device are used for respectively acquiring signals of gas-liquid two-phase flow under different working conditions, fluid fluctuation information is measured based on the double-helix high-frequency microwave sensor, fluid flow and density information is measured based on the precession vortex metering device, and the double-flow fusion network is provided for carrying out multi-sensor signal fusion and feature extraction and is used for measuring gas-liquid two-phase flow gas content under different working conditions.
The technical scheme adopted by the invention is as follows: a method for measuring the gas-liquid two-phase flow content, in particular to a method for measuring the gas-liquid two-phase flow content of an integrated multi-sensor. The method is characterized by comprising the following steps of:
(1) The double-helix high-frequency microwave sensor and the precession vortex metering device are constructed for measuring a gas-liquid two-phase flow water content fluctuation signal.
The double-helix high-frequency microwave sensor comprises a measuring pipe section, an excitation electrode, a receiving electrode, a protection electrode and a shielding layer, wherein the excitation electrode and the receiving electrode keep a wall-opposite type structure to rotate for 360 degrees at the outer side of a measuring pipeline, two protection electrodes which are synchronous to helix are arranged between the excitation electrode and the receiving electrode, and the shielding layer is arranged outside the electrodes to prevent an electric field from diffusing to the edge, so that the effects of protecting the electric field and enhancing the intensity of the electric field inside the pipeline are achieved.
The precession vortex metering device comprises a measuring pipe section, a throttling device, a differential pressure flowmeter and a precession vortex flowmeter, wherein the throttling device and the precession vortex flowmeter are arranged on a measuring pipeline, and the throttling device is connected with the differential pressure flowmeter.
(2) Carrying out dynamic experiments to obtain multi-sensor signals, specifically measuring gas-liquid two-phase flow fluctuation signals by adopting a double-helix high-frequency microwave sensor, and converting the gas-liquid two-phase flow fluctuation signals into microwave difference frequency signals; simultaneously, a precession vortex metering device is adopted to acquire flow and density signals;
when the gas-liquid two-phase flow flows through the measuring pipe section of the double-helix high-frequency microwave sensor, a sine excitation signal source generates a high-frequency excitation signal, one path of the high-frequency excitation signal is sent to an excitation electrode of the helical high-frequency microwave sensor through a power divider, the other path of the high-frequency excitation signal is sent to a monitor, the excitation electrode transmits high-frequency electromagnetic waves into the gas-liquid two-phase flow in a pipeline, the absorbed wave energy is different due to the fact that the content of polar water molecules in the gas-liquid two-phase flow is different under different working conditions, the induction electrode transmits different signals caused by the content difference of the polar water molecules to the monitor, the two paths of signals are subjected to signal mixing in the monitor, and one path of microwave difference frequency signal is obtained after the signals are processed by an adder.
When the gas-liquid two-phase flow flows through the measuring pipe section of the precession vortex metering device, the total flow Q of the fluid in the measuring pipe is obtained based on the throttling device and the differential pressure flowmeter 1 ：
Where u is the flow coefficient of the differential pressure flowmeter, S 0 Where ρ is the fluid density and ΔP is the differential pressure. Precession frequency f for measuring pressure in pipeline and total flow Q for measuring fluid in pipeline based on precession vortex flowmeter 2 ：
K in the formula x For the flow coefficient of the precession vortex flowmeter, the flow Q measured by the differential pressure flowmeter and the precession vortex flowmeter arranged on the same pipeline 1 And Q 2 The same way of density signal ρ is obtained through operation, and the formula is as follows:
(3) The method comprises the steps of constructing a data set, specifically preprocessing data, then acquiring samples from the data by using a sliding window, adding corresponding labels for the samples, and taking 80% of the samples as a training set and 20% of the samples as a testing set. The method specifically comprises the following steps:
the microwave difference frequency signal and the density signal are respectively preprocessed according to the following formula:
wherein I is o Represents the o data, I mean And I std The mean and standard deviation of the data are respectively,is the o data after pretreatment; the preprocessed microwave difference frequency signal and the preprocessed microwave density signal are simultaneously segmented by adopting a sliding window, the window length is H, and data with the length of P are acquired +.>Samples of (a) wherein->The representative is rounded downwards, the actual air content value is taken as a data label, and N samples with label values are obtained; randomly dividing N fluctuation samples into data sets with training sets and test sets, wherein the specific proportion is [ training set: test set ]]＝[8∶2]。
(4) The double-flow fusion network is provided for carrying out multi-sensor signal fusion and feature extraction and is used for measuring the gas-liquid two-phase flow gas content under different working conditions. The double-flow fusion network comprises a parallel characteristic module, a characteristic fusion module and a parameter prediction module.
The parallel characteristic module consists of two identical branches, microwave difference frequency signals and density signals are input respectively, and characteristic extraction is carried out on the two paths of signals respectively. Each branch of the parallel feature module is a dual stream network structure, wherein one stream uses convolution kernels of larger size, and wherein one stream uses convolution kernels of smaller size, and features are extracted comprehensively from different scales or dimensions.
Each stream of the dual stream network architecture consists of three one-dimensional convolution modules, two max-pooling layers and one drop layer. Each one-dimensional convolution module consists of a convolution layer, a Batch Normalization layer and an activation function layer.
The convolution layer is used for extracting the time domain characteristics of the signals, and the convolution kernel used by the convolution layer is J k And k=1, 2 and 3 respectively correspond to three one-dimensional convolution modules, and the formula of the convolution layer is as follows:
wherein X is the input of the convolution layer, Y is the output characteristic of the convolution layer, f is the activation function of the convolution layer, u is the position index of the midpoint of the convolution kernel, G is the characteristic dimension of the input of the convolution layer, G is the characteristic index of the input of the time convolution layer, omega p，g，k And b p The p-th convolution kernel and offset in the convolution layer, L k For the size of the convolution kernel, S is the step size of the convolution kernel of the time convolution layer, in is the subscript shorthand input by the time convolution layer, and m and n are the position indexes of the extracted features.
The Batch Normalization layer is used for carrying out standardized treatment on the characteristics, so that the problem of unstable numerical values in the neural network is avoided, the characteristics of the same batch are distributed similarly, and the easy training property of the network is enhanced; the activation function layer adopts a Relu activation function.
The maximum pooling layer is used for selecting the maximum value in the pooling area as a representative value to realize feature dimension reduction and reduce calculation load. The discarding layer discards a part of neurons and connections thereof according to probability, so that the neural network model is forced to learn more global characteristics.
The calculation formula of the feature fusion module is as follows:
fe＝F 1 ||F 2
fe′＝f D (fe*w D +b D )
wherein F is 1 ，F 2 For the input feature to be fused, fe' is the output feature, f D To activate the function, w D ，b D Respectively the weight and the bias of the full connection layer; the parameter prediction module uses a Sigmoid activation function to realize the numerical measurement of the water content.
The super-parameters of the double-flow fusion network are obtained by updating weight parameters of the double-flow fusion network based on a training set by using an AMSGrad optimization algorithm, measuring errors between output and true content of the double-flow fusion network by using a Mean Square Error (MSE) as a loss function, and optimizing network parameters by using a minimized error as a criterion.
Due to the adoption of the technical scheme, the invention has the following advantages:
the invention provides a gas-liquid two-phase flow content measuring method integrating multiple sensors, which integrates multiple sensors, and comprises the steps of respectively collecting signals of gas-liquid two-phase flows under different working conditions by a double-helix high-frequency microwave sensor and a precession vortex measuring device, directly measuring flow, density and other information based on the precession vortex measuring device, fusing different sensor characteristic information, capturing flow information based on multiple angles of microwaves and precession vortex, and improving measuring precision, reliability and instantaneity.
The invention provides double-flow fusion based on multi-sensor signals for signal fusion and analysis, and features are comprehensively extracted from different scales or dimensions by using convolution kernels of different sizes, so that the measurement accuracy is improved.
Drawings
FIG. 1 is a schematic diagram of a double helix high frequency microwave sensor and precession vortex metering device;
fig. 2 is a block diagram of a dual stream converged network.
Detailed Description
The present invention will be described in detail with reference to the following examples and the accompanying drawings.
The invention discloses a multi-sensor integrated gas-liquid two-phase flow content measuring method, which integrates various sensors, and comprises the steps of respectively collecting signals of gas-liquid two-phase flows under different working conditions by a double-helix high-frequency microwave sensor and a precession vortex metering device, wherein the double-helix high-frequency microwave sensor is used for measuring fluid fluctuation information, the precession vortex metering device is used for measuring fluid flow, density and other information, and a double-flow fusion network is constructed for carrying out multi-sensor signal fusion and feature extraction so as to realize measurement of gas-liquid two-phase flow content under different working conditions.
The invention relates to a method for measuring the gas-liquid two-phase flow content of an integrated multi-sensor, which comprises the following steps:
(1) Constructing a double-helix high-frequency microwave sensor and a precession vortex metering device for measuring a gas-liquid two-phase flow water content fluctuation signal;
the double-helix high-frequency microwave sensor comprises a measuring pipe section, an excitation electrode, a receiving electrode, a protection electrode and a shielding layer, wherein the excitation electrode and the receiving electrode keep a wall-opposite type structure to rotate for 360 degrees at the outer side of a measuring pipeline, two protection electrodes which are synchronous to helix are arranged between the excitation electrode and the receiving electrode, and the shielding layer is arranged outside the electrodes to prevent an electric field from diffusing to the edge, so that the effects of protecting the electric field and enhancing the intensity of the electric field inside the pipeline are achieved.
The precession vortex metering device comprises a measuring pipe section, a throttling device, a differential pressure flowmeter and a precession vortex flowmeter, wherein the throttling device and the precession vortex flowmeter are arranged on a measuring pipeline, and the throttling device is connected with the differential pressure flowmeter.
(2) Carrying out dynamic experiments to obtain multi-sensor signals, specifically measuring gas-liquid two-phase flow fluctuation signals by adopting a double-helix high-frequency microwave sensor, and converting the gas-liquid two-phase flow fluctuation signals into microwave difference frequency signals; and meanwhile, a precession vortex metering device is adopted to acquire flow and density signals.
When the gas-liquid two-phase flow flows through the measuring pipe section of the double-helix high-frequency microwave sensor, a sine excitation signal source generates a high-frequency excitation signal, one path of the high-frequency excitation signal is sent to an excitation electrode of the helical high-frequency microwave sensor through a power divider, the other path of the high-frequency excitation signal is sent to a monitor, the excitation electrode transmits high-frequency electromagnetic waves into the gas-liquid two-phase flow in a pipeline, the absorbed wave energy is different due to the fact that the content of polar water molecules in the gas-liquid two-phase flow is different under different working conditions, the induction electrode transmits different signals caused by the content difference of the polar water molecules to the monitor, the two paths of signals are subjected to signal mixing in the monitor, and one path of microwave difference frequency signal is obtained after the signals are processed by an adder.
When the gas-liquid two-phase flow flows through the measuring pipe section of the precession vortex metering device, the total flow Q of the fluid in the measuring pipe is obtained based on the throttling device and the differential pressure flowmeter 1 ：
Where u is the flow coefficient of the differential pressure flowmeter, S 0 Where ρ is the fluid density and ΔP is the differential pressure. Precession frequency f for measuring pressure in pipeline and total flow Q for measuring fluid in pipeline based on precession vortex flowmeter 2 ：
K in the formula x For the flow coefficient of the precession vortex flowmeter, the flow Q measured by the differential pressure flowmeter and the precession vortex flowmeter arranged on the same pipeline 1 And Q 2 The same way of density signal ρ is obtained through operation, and the formula is as follows:
(3) The method comprises the steps of constructing a data set, specifically preprocessing the data, acquiring samples from the data by using a sliding window, adding corresponding labels for the samples, and taking 80% of the samples as a training set and 20% of the samples as a testing set. The method specifically comprises the following steps:
the microwave difference frequency signal and the density signal are respectively preprocessed according to the following formula:
wherein I is o Represents the o data, I mean And I std The mean and standard deviation of the data are respectively,is the o data after pretreatment; the preprocessed microwave difference frequency signal and the preprocessed microwave density signal are simultaneously segmented by adopting a sliding window, the window length is H, and data with the length of P are acquired +.>Samples of (a) wherein->The representative is rounded downwards, the actual air content value is taken as a data label, and N samples with label values are obtained; along with itThe machine divides N fluctuation samples into data sets with training sets and test sets, wherein the specific proportion is [ training sets: test sets ]]＝[8∶2]。
(4) The double-flow fusion network is provided for carrying out multi-sensor signal fusion and feature extraction and is used for measuring the gas-liquid two-phase flow gas content under different working conditions. The double-flow fusion network comprises a parallel characteristic module, a characteristic fusion module and a parameter prediction module.
The parallel characteristic module consists of two identical branches, microwave difference frequency signals and density signals are input respectively, and characteristic extraction is carried out on the two paths of signals respectively. Each branch of the parallel feature module is a dual stream network structure, wherein one stream uses convolution kernels of larger size, and wherein one stream uses convolution kernels of smaller size, and features are extracted comprehensively from different scales or dimensions.
Each stream of the dual stream network architecture consists of three one-dimensional convolution modules, two max-pooling layers and one drop layer. Each one-dimensional convolution module consists of a convolution layer, a Batch Normalization layer and an activation function layer.
The convolution layer is used for extracting the time domain characteristics of the signals, and the convolution kernel used by the convolution layer is J k And k=1, 2 and 3 respectively correspond to three one-dimensional convolution modules, and the formula of the convolution layer is as follows:
wherein x is the input of the convolution layer, Y is the output characteristic of the convolution layer, f is the activation function of the convolution layer, u is the position index of the midpoint of the convolution kernel, G is the characteristic dimension of the input of the convolution layer, G is the characteristic index of the input of the time convolution layer, omega p，g，k And b p The p-th convolution kernel and offset in the convolution layer, L k For the size of the convolution kernel, S is the step of the convolution kernel of the time convolution layerLong, in is subscript shorthand input by the time convolution layer, m, n are position indexes of the extracted features.
The Batch Normalization layer is used for carrying out standardized treatment on the characteristics, so that the problem of unstable numerical values in the neural network is avoided, the characteristics of the same batch are distributed similarly, and the easy training property of the network is enhanced; the activation function layer adopts a Relu activation function.
The maximum pooling layer is used for selecting the maximum value in the pooling area as a representative value to realize feature dimension reduction and reduce calculation load. The discarding layer discards a part of neurons and connections thereof according to probability, so that the neural network model is forced to learn more global characteristics.
The calculation formula of the feature fusion module is as follows:
fe＝F 1 ||F 2
fe′＝f D (fe*w D +b D )
wherein F is 1 ，F 2 For the input feature to be fused, fe' is the output feature, f D To activate the function, w D ，b D Respectively the weight and the bias of the full connection layer; the parameter prediction module uses a Sigmoid activation function to realize the numerical measurement of the water content.
The super-parameters of the double-flow fusion network are obtained by updating weight parameters of the double-flow fusion network based on a training set by using an AMSGrad optimization algorithm, measuring errors between output and true content of the double-flow fusion network by using a Mean Square Error (MSE) as a loss function, and optimizing network parameters by using a minimized error as a criterion.
Claims (6)
1. The method for measuring the gas-liquid two-phase flow content of the integrated multisensor is characterized by comprising the following steps of:
(1) Constructing a double-helix high-frequency microwave sensor and a precession vortex metering device for measuring a gas-liquid two-phase flow water content fluctuation signal;
(2) Carrying out dynamic experiments to obtain multi-sensor signals, specifically adopting a double-helix high-frequency microwave sensor to measure gas-liquid two-phase flow fluctuation signals, converting the gas-liquid two-phase flow fluctuation signals into microwave difference frequency signals, and simultaneously adopting a precession vortex metering device to obtain flow and density signals;
(3) Constructing a data set, namely preprocessing the data, acquiring samples from the data by using a sliding window, adding corresponding labels for the samples, and taking 80% of the samples as a training set and 20% of the samples as a test set;
(4) The double-flow fusion network is provided for carrying out multi-sensor signal fusion and feature extraction and is used for measuring the gas-liquid two-phase flow gas content under different working conditions.
2. The multi-sensor integrated gas-liquid two-phase flow content measurement method according to claim 1, wherein step (1) comprises:
the double-helix high-frequency microwave sensor comprises a measuring pipe section, an excitation electrode, a receiving electrode, a protection electrode and a shielding layer, wherein the excitation electrode and the receiving electrode keep a wall-opposite type structure to rotate for 360 degrees at the outer side of a measuring pipeline, two protection electrodes of synchronous helices are arranged between the excitation electrode and the receiving electrode, and the shielding layer is arranged outside the electrodes to prevent an electric field from diffusing to the edge, so that the effects of protecting the electric field and enhancing the intensity of the electric field in the pipeline are achieved;
the precession vortex metering device comprises a measuring pipe section, a throttling device, a differential pressure flowmeter and a precession vortex flowmeter, wherein the throttling device and the precession vortex flowmeter are arranged on a measuring pipeline, and the throttling device is connected with the differential pressure flowmeter.
3. The multi-sensor integrated gas-liquid two-phase flow content measurement method according to claim 1, wherein step (2) comprises:
when the gas-liquid two-phase flow flows through the measuring pipe section of the double-helix high-frequency microwave sensor, a sine excitation signal source generates a high-frequency excitation signal, one path of the high-frequency excitation signal is sent to an excitation electrode of the helical high-frequency microwave sensor through a power divider, the other path of the high-frequency excitation signal is sent to a monitor, the excitation electrode transmits high-frequency electromagnetic waves into the gas-liquid two-phase flow in a pipeline, the absorbed wave energy is different due to the fact that the content of polar water molecules in the gas-liquid two-phase flow is different under different working conditions, the induction electrode transmits different signals caused by the content difference of the polar water molecules to the monitor, the two paths of signals are subjected to signal mixing in the monitor, and one path of microwave difference frequency signals are obtained after the signals are processed by an adder;
when the gas-liquid two-phase flow flows through the measuring pipe section of the precession vortex metering device, the total flow Q of the fluid in the measuring pipe is obtained based on the throttling device and the differential pressure flowmeter 1 ：
Where u is the flow coefficient of the differential pressure flowmeter, S 0 For orifice area, ρ is fluid density, ΔP is differential pressure, and the precession frequency f of the pressure in the pipe and the total flow Q of the fluid in the pipe are measured based on the precession vortex flowmeter 2 ：
K in the formula x For the flow coefficient of the precession vortex flowmeter, the flow Q measured by the differential pressure flowmeter and the precession vortex flowmeter arranged on the same pipeline 1 And Q 2 The same way of density signal ρ is obtained through operation, and the formula is as follows:
4. the multi-sensor integrated gas-liquid two-phase flow content measurement method according to claim 1, wherein step (3) comprises:
the microwave difference frequency signal and the density signal are respectively preprocessed according to the following formula:
wherein I is o Represents the o data, I mean And I std The mean and standard deviation of the data are respectively,is the o data after pretreatment; the preprocessed microwave difference frequency signal and the preprocessed microwave density signal are simultaneously segmented by adopting a sliding window, the window length is H, and data with the length of P are acquired +.>Samples of (a) wherein->The representative is rounded downwards, the actual air content value is taken as a data label, and N samples with label values are obtained; the N fluctuation samples are randomly divided into data sets with training sets and test sets, and the specific proportion is [ training set: test set]＝[8：2]。
5. The method for measuring the gas-liquid two-phase flow content of the integrated multisensor according to claim 1, wherein the double-flow fusion network in the step (4) comprises a parallel feature module, a feature fusion module and a parameter prediction module;
the parallel characteristic module consists of two identical branches, the input is a microwave difference frequency signal and a density signal respectively, the characteristic extraction is carried out on the two paths of signals respectively, each branch of the parallel characteristic module is of a double-flow network structure, one of the branches uses a convolution kernel with a larger size, the other branch uses a convolution kernel with a smaller size, and the characteristics are comprehensively extracted from different scales or dimensions; each flow of the double-flow network structure consists of three one-dimensional convolution modules, two maximum pooling layers and a discarding layer, wherein each one-dimensional convolution module consists of a convolution layer, a Batch Normalization layer and an activation function layer; the convolution layer is used for extracting the time domain characteristics of the signals, and the convolution kernel used by the convolution layer is J k And k=1, 2 and 3 respectively correspond to three one-dimensional convolution modules, and the formula of the convolution layer is as follows:
wherein X is the input of the convolution layer, Y is the output characteristic of the convolution layer, f is the activation function of the convolution layer, u is the position index of the midpoint of the convolution kernel, G is the characteristic dimension of the input of the convolution layer, G is the characteristic index of the input of the time convolution layer, omega p,g,k And b p The p-th convolution kernel and offset in the convolution layer, L k S is the step length of the convolution kernel of the time convolution layer, in is the subscript shorthand input by the time convolution layer, and m and n are the position indexes of the extracted features; the Batch Normalization layer is used for carrying out standardized treatment on the characteristics, so that the problem of unstable numerical values in the neural network is avoided, the characteristics of the same batch are distributed similarly, and the easy training property of the network is enhanced; the activation function layer adopts a Relu activation function; the maximum pooling layer is used for selecting the maximum value in the pooling area as a representative value to realize characteristic dimension reduction and reduce calculation load; the discarding layer discards a part of neurons and the connection thereof according to probability, so that the neural network model is forced to learn more global characteristics;
the calculation formula of the feature fusion module is as follows:
fe＝F 1 ||F 2
fe′＝f D (fe*w D +b D )
wherein F is 1 ,F 2 For the input feature to be fused, fe' is the output feature, f D To activate the function, w D ,b D Respectively the weight and the bias of the full connection layer;
the parameter prediction module uses a Sigmoid activation function to realize the numerical measurement of the water content.
6. The method for measuring the gas-liquid two-phase flow content of the integrated multisensor according to claim 1, wherein the super-parameters of the double-flow fusion network are obtained by updating weight parameters of the double-flow fusion network based on a training set by using an AMSGrad optimization algorithm, measuring errors between output and true content of the double-flow fusion network by using a Mean Square Error (MSE) as a loss function and optimizing network parameters by using a minimized error as a criterion.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311628639.XA CN117664227A (en) | 2023-12-01 | 2023-12-01 | Multi-sensor integrated gas-liquid two-phase flow content measuring method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311628639.XA CN117664227A (en) | 2023-12-01 | 2023-12-01 | Multi-sensor integrated gas-liquid two-phase flow content measuring method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117664227A true CN117664227A (en) | 2024-03-08 |
Family
ID=90070775
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311628639.XA Pending CN117664227A (en) | 2023-12-01 | 2023-12-01 | Multi-sensor integrated gas-liquid two-phase flow content measuring method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117664227A (en) |
-
2023
- 2023-12-01 CN CN202311628639.XA patent/CN117664227A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Alanazi et al. | Application of neural network and time-domain feature extraction techniques for determining volumetric percentages and the type of two phase flow regimes independent of scale layer thickness | |
| Lei et al. | Underground cylindrical objects detection and diameter identification in GPR B-scans via the CNN-LSTM framework | |
| CN101984340A (en) | Method for identifying flow type of soft grain two-phase turbulence based on artificial intelligence | |
| CN111351540B (en) | Method and system for detecting mass flow rate of particles in pneumatic conveying process | |
| CN111460625B (en) | Method for constructing multi-phase flow online real-time measurement artificial intelligence model | |
| CN109283255B (en) | Detection method of conveying flow pattern in pneumatic conveying process | |
| CN113447879A (en) | Electric energy meter metering error prediction method, system and device based on multi-stress influence | |
| Bazan et al. | Multi-fault diagnosis in three-phase induction motors using data optimization and machine learning techniques | |
| Xu et al. | A cutting pattern recognition method for shearers based on improved ensemble empirical mode decomposition and a probabilistic neural network | |
| Zhang et al. | Amplitude‐frequency images‐based ConvNet: Applications of fault detection and diagnosis in chemical processes | |
| Huang et al. | Deep learning for magnetic flux leakage detection and evaluation of oil & gas pipelines: A review | |
| Martins et al. | Online monitoring of sensor calibration status to support condition-based maintenance | |
| CN110632132A (en) | High-yield gas-oil well wellhead liquid water content prediction method based on multi-sensor measurement and deep convolutional neural network | |
| Lee et al. | A robust deep neural network for rolling element fault diagnosis under various operating and noisy conditions | |
| CN117664227A (en) | Multi-sensor integrated gas-liquid two-phase flow content measuring method | |
| Tao et al. | Real-Time Fault Diagnosis for Hydraulic System Based on Multi-Sensor Convolutional Neural Network | |
| Ma et al. | High-sensitivity ultrasonic guided wave monitoring of pipe defects using adaptive principal component analysis | |
| Zhu et al. | Failure analysis and intelligent identification of critical friction pairs of an axial piston pump | |
| CN112700051A (en) | Res-TCN neural network-based intelligent prediction method for oil well liquid production associated gas | |
| CN110630244B (en) | High-yield gas-oil well water content prediction system and method based on multi-sensor measurement and long-time and short-time memory network | |
| CN110243886A (en) | A kind of low yield gas well mouth of oil well hydrated comples ion method based on nonlinear characteristic | |
| Guo et al. | Mechanical fault diagnosis of a dc motor utilizing united variational mode decomposition, sampen, and random forest-sprint algorithm classifiers | |
| CN113281409A (en) | Method and system for detecting solid flow form and concentration in hydraulic conveying process | |
| Hou et al. | A fast method for Chinese liquor recognition | |
| Zhao et al. | Research on Transformer Oil Multi-frequency Ultrasonic Monitoring Technology Based on Convolutional Neural Network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |