CN110632132A - High-yield gas-oil well wellhead liquid water content prediction method based on multi-sensor measurement and deep convolutional neural network - Google Patents
High-yield gas-oil well wellhead liquid water content prediction method based on multi-sensor measurement and deep convolutional neural network Download PDFInfo
- Publication number
- CN110632132A CN110632132A CN201910613946.8A CN201910613946A CN110632132A CN 110632132 A CN110632132 A CN 110632132A CN 201910613946 A CN201910613946 A CN 201910613946A CN 110632132 A CN110632132 A CN 110632132A
- Authority
- CN
- China
- Prior art keywords
- water content
- sensor
- frequency
- time
- wellhead
- 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
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N27/00—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means
- G01N27/02—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating impedance
- G01N27/22—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating impedance by investigating capacitance
- G01N27/223—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating impedance by investigating capacitance for determining moisture content, e.g. humidity
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
- G01N29/02—Analysing fluids
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
- G01N29/44—Processing the detected response signal, e.g. electronic circuits specially adapted therefor
- G01N29/4481—Neural networks
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2291/00—Indexing codes associated with group G01N29/00
- G01N2291/02—Indexing codes associated with the analysed material
- G01N2291/022—Liquids
- G01N2291/0226—Oils, e.g. engine oils
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2291/00—Indexing codes associated with group G01N29/00
- G01N2291/02—Indexing codes associated with the analysed material
- G01N2291/024—Mixtures
- G01N2291/02433—Gases in liquids, e.g. bubbles, foams
Abstract
The invention relates to a high-yield gas-oil well wellhead water content prediction method based on multi-sensor measurement and a convolutional neural network. In the aspect of measurement, the water content information of three-phase flow of liquid produced by a wellhead is measured through a high-frequency dual-ring type capacitive sensor, the gas content information of the three-phase flow of the liquid produced by the wellhead is obtained through an ultrasonic transmission type sensor, a water content time-frequency spectrogram and a gas content time-frequency spectrogram are respectively extracted and used as model training characteristic input, a deep convolution neural network is adopted to perform characteristic fusion on the water content time-frequency spectrogram and the gas content time-frequency spectrogram and abstract high-dimensional water content characteristic expression layer by layer, and a water content test value of the wellhead is used as a water content label to train a water content prediction model of a. Because the time-frequency characteristics of the water content fluctuation signal and the gas content fluctuation signal are accurate description of the characteristics of the liquid produced by the wellhead, the method can effectively improve the measurement accuracy of the water content of the liquid produced by the wellhead.
Description
Technical Field
The invention relates to a method for measuring the water content of liquid produced by a high-yield gas-oil well, in particular to a method for predicting the water content of liquid produced by a well head of the high-yield gas-oil well based on multi-sensor measurement and a deep convolutional neural network
Background
In the process of crude oil production, timely grasping and controlling the water content parameters of the produced liquid of the oil well not only is a premise of reliably estimating the net yield of crude oil, but also is a basis for correctly diagnosing and maintaining problems of the oil well, and the water content parameters of the produced liquid of the oil well are also important guide indexes for adjusting the exploitation mode of an oil reservoir, so that the detection of the water content parameters of the produced liquid of the oil well has important significance. At present, the ultrahigh water content characteristic of oil field produced fluid puts new requirements on the measurement of the water content of oil well produced fluid, but how to accurately acquire the water content information of the high water content oil well produced fluid becomes a problem to be solved urgently. Currently, the detection of the water content of oil well production liquid is usually realized by a specially designed sensor, and the measuring method comprises an ultrasonic method, an optical method, a ray method, an imaging method, an electric conduction method, an electric method and the like. However, the measurement effect of the existing sensor can not meet the requirement under the working condition of high water content of oil well produced liquid, and is represented by the fact that the response nonlinearity and the water content resolution ratio of the sensor are low, and the measurement result is greatly influenced by the mineralization degree. In addition, the traditional assay method in the oilfield operation is greatly influenced by the sampling condition and the sampling frequency, the measurement period is long, and the real-time measurement is difficult to realize. Due to the fact that the gas content of the high-yield gas-oil well is too large, the high gas content has great influence on the prediction of the water content.
Although the soft measurement of the water content of the oil-water two-phase flow through a neural network or a shallow network such as a support vector machine is widely applied, the characteristics of the shallow network structure need to be carefully designed in the application process. Generally, the shallow feature has strong subjectivity, and the prediction result of the model on the water content is also greatly influenced by the designed feature. In recent years, the artificial intelligence technology is widely applied in the industrial field, and especially the application of the deep learning method widens the application range of the artificial intelligence technology. The deep learning technology is a new theory which is emerging in recent years, the characteristics of the measured object are extracted layer by layer in an unsupervised mode, the objectivity of the characteristics is strong, and the essence of the measured object can be accurately and accurately reflected.
Through the search of the patent publications, two patent publications similar to the purpose and technical scheme of the patent application are found:
1. a method (109447342A) for predicting initial water cut of an ultra-low permeability sandstone reservoir oil well during production, the method comprising: collecting, sorting and selecting ultra-low permeability sandstone oil reservoir calculation parameters; and predicting the initial water content of the ultra-low permeability sandstone reservoir oil well in the production period by using the functional relation between the effective stress and the water saturation. The method for predicting the initial water content of the ultra-low permeability sandstone reservoir oil well in the production period provides theoretical basis for explaining and revealing the initial water content of the oil reservoir oil well in the production period and predicting the initial water content of the oil well in the production period, and achieves the purpose of dynamically predicting the initial water content of the ultra-low permeability sandstone reservoir oil well in the production period, so that the method has certain theoretical and practical significance.
2. A multi-model prediction method (105631554A) for oil water content of an oil well based on time series is characterized by comprising the following steps: 1) establishing an oil well oil water content data set of { xi, i ═ 1,2, …, N } by using historical data; 2) preprocessing data in an oil water content data set { xi, i ═ 1,2, …, N } by adopting a wavelet analysis method; 3) classifying the { xi } Wave by a neighbor propagation clustering algorithm; 4) and representing the data in each cluster by the following time series form: 5) and establishing a time series model of each cluster according to an extreme learning machine algorithm and obtaining a predicted value by using the time series model. It has solved the artifical sample of current oil fluid moisture content wastes time and energy, influences the problem of production control and oil recovery data's real-time.
Through comparison of technical characteristics, the adopted oil reservoir calculation parameters and modes in the comparison document 1 are fundamentally different from those in the application of the invention; although the comparison document 2 adopts a time series method to predict the water content, the water content model and the method thereof are fundamentally different from those of the present invention, and thus, the present invention is not substantially creative.
Disclosure of Invention
The invention aims to solve the technical problem of providing a method for predicting the water content of the liquid produced by a wellhead of a high-yield gas-oil well based on multi-sensor measurement and a deep convolutional neural network, and the method can realize accurate prediction of the water content of the liquid produced by an oil well, particularly the liquid produced by the high-yield gas-oil well, by fusing the gas content fluctuation and the water content fluctuation of three-phase flow at the wellhead through a water content prediction model.
The technical scheme for realizing the purpose of the invention is as follows:
a method for predicting the water content of liquid produced by a wellhead of a high-yield gas-oil well based on multi-sensor measurement and a deep convolutional neural network is characterized by comprising the following steps of: the method comprises the following steps:
(1) obtaining water content information and gas content information of three-phase flow of a wellhead
The sensor assembly comprises a double-ring type capacitance sensor and an ultrasonic sensor, the double-ring type capacitance sensor is used for acquiring water content information of a wellhead, and a capacitance sensor measuring signal is obtained through a sensor measuring circuit of the water content multi-element time sequence characteristic extraction module; the ultrasonic sensor is used for acquiring wellhead gas content information, exciting a transmitting probe of the ultrasonic sensor by an ultrasonic excitation signal, and measuring a signal of an ultrasonic receiving probe to obtain a measuring signal of the ultrasonic sensor;
(2) preprocessing of multi-sensor acquired signals
Performing Hilbert transform on measurement signals acquired by the two sensors, and then respectively calculating WVD distribution and filtering to obtain a capacitance signal time-frequency characteristic diagram and an ultrasonic signal time-frequency characteristic diagram which are used as training characteristics of an intelligent prediction model of the water content of the high-yield gas well;
defining the time sequence of the water content measured by the sensor as x (t), processing the x (t) through Hilbert transform to obtain an analytic form Z (t) and a conjugate analytic form Z (t), and then calculating the WVD distribution of a water content time sequence fluctuation signal:
wherein f is frequency, t is time, τ is time delay, and z (t) is the analytic form of the original signal;
in order to eliminate cross terms in the WVD time-frequency distribution plane, the WVD is filtered through a filter function, and the cross terms in the time-frequency distribution are eliminated. The time-frequency distribution of the obtained water content fluctuation information is as follows:
P(t,f)＝∫∫φ(t,υ)·WVD(t-τ,f-υ)dτdυ
p (t, f) is a time-frequency joint distribution characteristic of the water content time sequence information obtained through calculation, tau is time delay, upsilon is frequency offset, phi (t, upsilon) is exp (-4 pi)2υ2τ2/σ) is a filter kernel function;
(3) forecasting the water content of the wellhead of the high-yield gas-oil well through the time-frequency characteristics of the deep convolution neural network learning water content and gas content signals
Overlapping and fusing the time-frequency joint distribution maps of signals acquired by the double-ring high-frequency capacitance sensor and the transmission type ultrasonic sensor to obtain a fused time-frequency joint distribution map which is used as the input of a network for training; and extracting the original time-frequency joint characteristics layer by adopting a deep convolutional neural network structure, and refining highly abstract moisture content characteristic information.
In the step (1), the dual-ring capacitance sensor measures the wellhead content by adopting a continuous measurement mode, the sampling frequency is set to be 10 times per minute, and the measurement data is a typical time sequence of reaction content change.
In the step (1), the capacitance sensor measurement signal is acquired by: the structure of the sensor measuring circuit is as follows: the high-frequency sine excitation signal source generates an excitation signal, the excitation signal is sent to an annular measuring electrode of the double-ring type capacitance sensor through the power divider to sweep frequency, the annular measuring electrode excites water content data measured by the sweep frequency and then enters the frequency mixer to carry out signal frequency mixing, and the signal after the frequency mixing is subjected to voltage bias through the adder to obtain a capacitance sensor measuring signal.
In step (3), the deep convolutional neural network includes 5 convolutional layers, and the first layer, the second layer, and the fifth convolutional layer are pooled to prevent the over-fitting phenomenon. The first layer of convolution operation adopts 48 convolution kernels with the size of 11 x 11, the step size is set to be 4, and then pooling operation is carried out, the size of the pooling convolution kernel is 3 x 3, and the step size is 2; the second layer of convolution adopts 128 convolution kernels with the size of 5 x 5, the step size is set to be 1, the size of the pooling convolution kernel is 3 x 3, and the step size is 2; then, the third layer and the fourth layer are convolution layers, the pooling operation is not carried out, the sizes of convolution kernels are all set to be 3 x 3, and the number of convolution kernels is all set to be 192; the fifth layer is a convolution layer, 128 convolution kernels with the size of 3 x 3 are arranged, the step size is 1, the size of the pooling convolution kernel is 3 x 3, and the step size is 2; features extracted by the deep convolutional neural network are input into a two-layer full-connection network, the number of nodes of each layer of the full-connection network is 1024, and finally, the water content high-abstraction features are input into a softmax classifier to predict the water content.
The two ends of the stainless steel protective shell of the sensor are provided with a left flange and a right flange, wherein the right flange is connected with the stainless steel protective shell in a threaded engagement mode; left flange and right flange and well head descent pipe connection, the inside coaxial arrangement of stainless steel protective housing has the woollen goods pipeline, interval installation has two annular sensor measuring electrode on woollen goods pipeline outer wall, coaxial arrangement has the electromagnetic shield layer in the sensor measuring electrode outside, the woollen goods pipeline compresses tightly sealedly with the stainless steel protective housing through the O type circle on top, still install transmission-type ultrasonic sensor on woollen goods pipeline outer wall, open the stainless steel protective housing lateral wall has the pin hole, organic glass ring support between electromagnetic shield layer and the woollen goods pipeline.
The invention has the advantages and positive effects that:
1. aiming at the problems that the gas content in a high-gas-yield oil well is too large and the prediction of the water content is greatly influenced, the double-ring type capacitance sensor is adopted to quickly and accurately obtain a water content sequence fluctuation signal; the adopted ultrasonic sensor is sensitive to gas enough, and can accurately measure gas fraction information in multiphase flow; moreover, by windowing the signals, the water content multivariate characteristic sequences of different time periods can be extracted, then the weighted summation is carried out on the segmentation signals, and the water content sequences and the gas content sequences can be fused, so that partial information is not lost, and the overall sequence characteristics can be obtained; analyzing the fused time sequence characteristics in a time-frequency domain and a recursion domain to obtain a multi-element characteristic value, and highlighting the multi-dimensional characteristics of the signals; the training of the depth long-short time memory (LSTM) neural network on the multidimensional characteristic sequence can accurately predict the water content value of the wellhead, realize the accurate measurement of the water content of the wellhead produced liquid of the high gas production oil well and simultaneously reduce the influence of the wellhead produced gas on the measurement.
2. The multi-sensor is arranged on a wellhead descending pipeline, the water content and the gas content in the pipeline can be directly measured, the measured measurement value can truly reflect the liquid production condition of a measured oil well, and the method has important significance for guiding the optimization management of the oil field; compared with the existing sensor, the sensor has stronger stability, the shielding layer can effectively shield the scattering of microwaves and the interference of external electromagnetic waves, the signal is locked in a certain range, the gas-liquid flow condition in the low-gas-production oil well pipeline can be effectively and accurately measured, the water content and the gas content can be simultaneously obtained through the measurement of the two sensors, and a new idea is provided for multiphase flow measurement and data fusion.
3. According to the method, the time-frequency joint distribution characteristics of the time sequence signals measured by the sensor are extracted, the characteristics are stored in a characteristic matrix form, and the characteristics contain abundant wellhead water content and gas content fluctuation information; through the deep learning of the characteristics, the basic characteristics and the rules of the water content change can be captured, and a data basis is provided for a wellhead water content prediction model based on artificial intelligence.
4. In order to realize accurate measurement of the water content of the liquid produced by the wellhead of the high-gas-production oil well and reduce the influence of gas production of the wellhead on the measurement, in the measurement, a high-frequency capacitive sensor is used for acquiring a content fluctuation signal of a mixed liquid, a transmission-type ultrasonic sensor is used for acquiring a content fluctuation signal in a multiphase flow, a time-frequency characteristic diagram of the measurement signal is extracted to be used as an input characteristic of a deep convolution neural network, a water content prediction model based on the deep convolution neural network is established, a water content test value of the liquid produced by the wellhead is used as a water content label for training, and finally a water content prediction value is. The method can effectively eliminate the influence of a small amount of gas at the wellhead on the measurement, and further improve the measurement precision of the water content of the liquid produced at the wellhead.
5. The deep convolutional neural network has objectivity in predicting the water content, eliminates uncertainty and subjectivity of manual operation in testing, and has stronger prediction performance and objectivity on data after training of a large amount of data and iteration times; meanwhile, compared with the traditional algorithm, such as a support vector machine, the deep convolutional neural network has better prediction accuracy, and the prediction accuracy can reach more than 98%.
Drawings
FIG. 1 is a structural cross-sectional view of a dual-ring high-frequency capacitance sensor for measuring the water content and gas content of three-phase flow at a wellhead according to the invention;
FIG. 2 is a schematic diagram of a time-frequency feature extraction module for water content and gas content of a wellhead according to the present invention;
FIG. 2-1 is an enlarged view of the capacitive sensor measurement signal of FIG. 2;
FIG. 2-2 is an enlarged view of the ultrasonic sensor measurement signal of FIG. 2;
FIG. 3 is a flow chart of a high-yield gas-oil well wellhead water content prediction method based on a deep convolutional neural network.
Detailed Description
The invention is further illustrated by the following examples: the following examples are illustrative and not intended to be limiting, and are not intended to limit the scope of the invention.
According to the method, the water content and gas content information of the wellhead is acquired through a double-ring type capacitance sensor and a transmission type ultrasonic sensor of a sensor assembly, time-frequency characteristics of a measurement signal are extracted by a water content multi-element time sequence characteristic extraction module to serve as the input of a deep convolution neural network, the network abstracts, extracts and synthesizes the input time-frequency joint distribution characteristics, and a wellhead water content prediction model is obtained by adopting a supervised learning mode. The structures of the components are described below:
1. the sensor assembly comprises a double-ring type capacitance sensor and an ultrasonic sensor, and the two sensors are used for acquiring the water content and gas content information of the wellhead. The structural cross-section of the sensor assembly is shown in fig. 1. The two ends of the stainless steel protective shell 3 of the sensor are a left flange 1 and a right flange 11 with nominal diameter DN50, wherein the right flange is connected with the stainless steel protective shell in a threaded 10 meshing mode, so that the internal components of the stainless steel protective shell can be conveniently installed. The left flange and the right flange are connected with a wellhead descending pipeline. The inside coaxial arrangement of stainless steel protective housing has the woollen goods pipeline 8 of the woollen goods material that the internal diameter is 50mm for the transmission of well head oil water mixed liquid. Two annular sensor measuring electrodes 5 are installed on the outer wall of the woolen cloth pipe at intervals and used for obtaining the water content information of the oil-gas-water three-phase mixed liquid. And an electromagnetic shielding layer 4 is coaxially arranged outside the measuring electrode so as to improve the measuring effect of the sensor. The woolen cloth pipeline is tightly pressed and sealed with the stainless steel protective shell through the O-shaped ring 2 at the top end, so that leakage of liquid produced at the wellhead is prevented. In addition, a transmission type ultrasonic sensor 9 is also arranged on the outer wall of the woolen cloth pipeline and used for acquiring the gas content information of the oil-gas-water three-phase mixed liquid. The side wall of the stainless steel protective shell is provided with a lead hole 6 with the inner diameter of 18mm, and the lead hole is used for connecting a measuring electrode of the sensor with a connecting wire channel of an external measuring and calculating instrument. In this embodiment, sensor stainless steel metal protection shell flange interval is 330mm, the woollen goods pipeline length in the sensor protection shell is 310mm, the pipeline latus rectum is 50mm, woollen goods pipeline wall thickness 80mm, cyclic annular sensor measuring electrode internal diameter 80mm, external diameter 85mm, two electrode spacing 20mm, the shielding layer is 1 mm's metal copper plate for thickness, the roll welding is a cylinder section of thick bamboo, length is 40mm, the internal diameter is 90mm, and woollen goods pipeline between organic glass ring 7 support.
2. The structure of the water content multi-element time sequence feature extraction module is shown in figure 2. Wherein, the left side of fig. 2 is a schematic diagram of a sensor measuring circuit, a high-frequency sinusoidal excitation signal source generates an excitation signal, the excitation signal is sent to a ring-shaped sensor measuring electrode through a power divider to carry out frequency sweeping, the ring-shaped sensor measuring electrode excites water content data measured by the frequency sweeping and then enters a frequency mixer to carry out signal frequency mixing, and the signal after frequency mixing is subjected to an adder and voltage bias to obtain a capacitance sensor measuring signal; the ultrasonic excitation signal excites the transmitting probe of the ultrasonic sensor, and the signal of the ultrasonic receiving probe is measured to obtain the measuring signal of the ultrasonic sensor. Fig. 2 is a schematic diagram of time-frequency characteristic analysis on the right side, in which time sequences of a capacitance sensor measurement signal and an ultrasonic sensor measurement signal are set as x (t), x (t) is processed through hilbert transform to obtain an analytic form Z (t) and a conjugate analytic form Z × t, and then WVD distribution of water content and gas content time sequence fluctuation signals is calculated:
where f is the frequency, t is the time, τ is the time delay, and z (t) is the analytic form of the original signal.
In order to eliminate cross terms in the WVD time-frequency distribution plane, the WVD is filtered through a filter function, and the cross terms in the time-frequency distribution are eliminated. The time-frequency distribution of the obtained water content fluctuation information is as follows:
P(t,f)＝∫∫φ(t,υ)·WVD(t-τ,f-υ)dτdυ
p (t, f) is the time-frequency joint distribution characteristic of the water content and gas content time sequence information obtained through calculation, tau is time delay, upsilon is frequency offset, phi (t, upsilon) is exp (-4 pi)2υ2τ2/σ) is the filter kernel function. And obtaining a capacitance signal time-frequency characteristic graph and an ultrasonic signal time-frequency characteristic graph through the calculation.
3. The structure of the deep convolutional neural network is shown in fig. 3. And fusing the capacitance signal time-frequency characteristic graph and the ultrasonic signal time-frequency characteristic graph in an overlapping way to obtain a fused sensor measurement time-frequency distribution graph, and transmitting the frequency distribution graph serving as input data of the deep convolution neural network into a convolution layer pooling layer. The invention adopts 5 layers of convolution Conv, wherein Pooling operation is used for one, two and five layers, and non-Pooling operation is used for the other three and four layers; the convolutional layer can upgrade and refine the features in the time-frequency distribution diagram, the convolutional layer output is transmitted into a full-connection layer, the full-connection layer has 2 layers, each layer comprises 1024 neurons, and the water content depth features are obtained; the neuron transmits the output to a Softmax classification function for prediction, a deep convolutional neural network is supervised learning, so that a well head water content test value is adopted as a water content label value, the predicted value and the label value are distinguished through the Softmax classification function, reverse correction of network parameters is carried out until the maximum iteration number reaches 10,000 steps, and training is completed; the trained network can accurately predict the moisture content time-frequency distribution diagram, and a predicted value of the moisture content is output through Softmax.
A method for predicting the water content of liquid produced by a wellhead of a high-yield gas-oil well based on multi-sensor measurement and a deep convolutional neural network comprises the following steps:
(1) obtaining water content information and gas content information of three-phase flow of a wellhead
The sensor assembly is installed in the well head descending pipeline, and is connected into the pipeline through DN50 flange connection. According to the invention, the double-ring type capacitance sensor of the sensor assembly is adopted to acquire the water content information of the wellhead three-phase flow, and the frequency sweeping operation is carried out on the double-ring type capacitance sensor so as to determine the optimal working frequency of the sensor. The sweep frequency band of the sensor is set to be 0.8Ghz-10GHz, which is a microwave band. After the optimal working frequency of the sensor is determined, exciting the measuring electrode of the sensor at the frequency, and measuring amplitude attenuation and phase attenuation of the microwave signal after passing through the sensor to serve as raw measuring information of the water content. The dual-ring high-frequency capacitance sensor adopts a continuous measurement mode for measuring the wellhead content, the sampling frequency is set to be 10 times per minute, and the measurement data is a typical time sequence of reaction content change; the method comprises the steps of acquiring gas content information of three-phase flow of a wellhead by adopting a transmission type ultrasonic sensor, wherein the ultrasonic sensor is provided with two probes, one probe is used as a transmitting probe, the other probe is used as a receiving probe, ultrasonic excitation is carried out on the transmitting probe, a time sequence which reflects typical gas content change is obtained by measuring signals of the receiving probe, and a time sequence value measured by the sensor can be uploaded to a server in a wireless transmission mode to be stored and analyzed.
(2) Preprocessing of multi-sensor acquired signals
And performing Hilbert transform on the measurement signals acquired by the sensor, and then respectively distributing and filtering by using a calculator WVD (WVD) to obtain a capacitance signal time-frequency characteristic diagram and an ultrasonic signal time-frequency characteristic diagram.
Defining the time sequence of water content measured by the sensor as x (t), and processing the x (t) through Hilbert transform to obtain an analytic form Z (t) and a conjugate analytic form Z*(t) of (d). And then calculating the WVD distribution of the water content time sequence fluctuation signal:
wherein f is frequency, t is time, τ is time delay, and z (t) is the analytic form of the original signal;
in order to eliminate cross terms in the WVD time-frequency distribution plane, the WVD is filtered through a filter function, and the cross terms in the time-frequency distribution are eliminated. The time-frequency distribution of the obtained water content fluctuation information is as follows:
P(t,f)＝∫∫φ(t,υ)·WVD(t-τ,f-υ)dτdυ
p (t, f) is a time-frequency joint distribution characteristic of the water content time sequence information obtained through calculation, tau is time delay, upsilon is frequency offset, phi (t, upsilon) is exp (-4 pi)2υ2τ2/σ) is the filter kernel function.
According to the method, the time-frequency joint distribution characteristics of the time sequence signals measured by the sensor are extracted, the characteristics are stored in a characteristic matrix form, and the characteristics contain abundant wellhead water content and gas content fluctuation information; through the deep learning of the characteristics, the basic characteristics and the rules of the water content change can be captured, and a data basis is provided for a wellhead water content prediction model based on artificial intelligence.
(3) Establishment of water content model based on artificial intelligence
And predicting the water content of the wellhead of the high-yield gas-oil well by learning the time-frequency characteristics of the water content and the gas content signals through a deep convolutional neural network. Overlapping and fusing time-frequency joint distribution maps of signals acquired by a double-ring high-frequency capacitance sensor and a transmission-type ultrasonic sensor to obtain a fused time-frequency joint distribution map, training the fused time-frequency joint distribution map as the input of a network, extracting original time-frequency joint features layer by adopting a deep convolutional neural network structure, and refining highly abstract water content feature information; in this embodiment, the deep convolutional neural network includes 5 convolutional layers, wherein the first convolutional layer, the second convolutional layer, and the fifth convolutional layer are all performed with pooling operation to prevent the over-fitting phenomenon. The first layer of convolution operation adopts 48 convolution kernels with the size of 11 x 11, the step size is set to be 4, and then pooling operation is carried out, the size of the pooling convolution kernel is 3 x 3, and the step size is 2; the second layer of convolution adopts 128 convolution kernels with the size of 5 x 5, the step size is set to be 1, the size of the pooling convolution kernel is 3 x 3, and the step size is 2; then, the third layer and the fourth layer are convolution layers, the pooling operation is not carried out, the sizes of convolution kernels are all set to be 3 x 3, and the number of convolution kernels is all set to be 192; the fifth layer is a convolution layer, 128 convolution kernels with the size of 3 x 3 are arranged, the step size is 1, the size of the pooling convolution kernel is 3 x 3, and the step size is 2; the characteristics extracted by the deep convolutional neural network are input into a two-layer fully-connected network, the number of nodes of each layer of the fully-connected network is 1024, finally, the high-abstract characteristics of the water content are input into a softmax classifier to predict the water content, the establishment of the water content prediction model in the embodiment belongs to a supervised learning mode, and the tag value of the water content is derived from the water content test value of the oil well wellhead produced liquid. And (4) carrying out positive and false judgment on the predicted value and the tag value through a softmax function, and reversely transmitting a judgment parameter into the deep convolutional neural network to carry out parameter correction until the network training is finished. The number of training iterations was set to 10,000 steps, with a batch size of 100. The trained network can accurately predict the time-frequency distribution diagram of the water content fluctuation curve, so that the water content fluctuation curve can be predicted.
Claims (5)
1. A method for predicting the water content of liquid produced by a wellhead of a high-yield gas-oil well based on multi-sensor measurement and a deep convolutional neural network is characterized by comprising the following steps of: the method comprises the following steps:
(1) obtaining water content information and gas content information of three-phase flow of a wellhead
The sensor assembly comprises a double-ring type capacitance sensor and an ultrasonic sensor, the double-ring type capacitance sensor is used for acquiring water content information of a wellhead, and a capacitance sensor measuring signal is obtained through a sensor measuring circuit of the water content multi-element time sequence characteristic extraction module; the ultrasonic sensor is used for acquiring wellhead gas content information, exciting a transmitting probe of the ultrasonic sensor by an ultrasonic excitation signal, and measuring a signal of an ultrasonic receiving probe to obtain a measuring signal of the ultrasonic sensor;
(2) preprocessing of multi-sensor acquired signals
Performing Hilbert transform on measurement signals acquired by the two sensors, and then respectively calculating WVD distribution and filtering to obtain a capacitance signal time-frequency characteristic diagram and an ultrasonic signal time-frequency characteristic diagram which are used as training characteristics of an intelligent prediction model of the water content of the high-yield gas well;
defining the time sequence of water content measured by the sensor as x (t), and processing the x (t) through Hilbert transform to obtain an analytic form Z (t) and a conjugate analytic form Z*(t), then calculating the WVD distribution of the water cut rate time series fluctuation signal:
wherein f is frequency, t is time, τ is time delay, and z (t) is the analytic form of the original signal;
in order to eliminate cross terms in a WVD time-frequency distribution plane, the WVD is filtered through a filtering function, the cross terms in the time-frequency distribution are eliminated, and the time-frequency distribution of the water content fluctuation information is obtained as follows:
P(t,f)＝∫∫φ(t,υ)·WVD(t-τ,f-υ)dτdυ
p (t, f) is a time-frequency joint distribution characteristic of the water content time sequence information obtained through calculation, tau is time delay, upsilon is frequency offset, phi (t, upsilon) is exp (-4 pi)2υ2τ2/σ) is a filter kernel function;
(3) forecasting the water content of the wellhead of the high-yield gas-oil well through the time-frequency characteristics of the deep convolution neural network learning water content and gas content signals
Overlapping and fusing the time-frequency joint distribution maps of signals acquired by the double-ring high-frequency capacitance sensor and the transmission type ultrasonic sensor to obtain a fused time-frequency joint distribution map which is used as the input of a network for training; and extracting the original time-frequency joint characteristics layer by adopting a deep convolutional neural network structure, and refining highly abstract moisture content characteristic information.
2. The method for predicting the water content of the liquid produced by the wellhead of the high-yield gas-oil well based on the multi-sensor measurement and the deep convolutional neural network as claimed in claim 1, wherein the method comprises the following steps: in the step (1), the dual-ring type capacitance sensor adopts a continuous measurement mode for measuring the wellhead occupancy rate, the sampling frequency is set to be 10 times per minute, and the measurement data is a typical time sequence of reaction occupancy rate change.
3. The method for predicting the water content of the liquid produced by the wellhead of the high-yield gas-oil well based on the multi-sensor measurement and the deep convolutional neural network as claimed in claim 1, wherein the method comprises the following steps: in the step (1), the capacitive sensor measurement signal is obtained by: the structure of the sensor measuring circuit is as follows: the high-frequency sine excitation signal source generates an excitation signal, the excitation signal is sent to an annular measuring electrode of the double-ring type capacitance sensor through the power divider to sweep frequency, the annular measuring electrode excites water content data measured by the sweep frequency and then enters the frequency mixer to carry out signal frequency mixing, and the signal after the frequency mixing is subjected to voltage bias through the adder to obtain a capacitance sensor measuring signal.
4. The method for predicting the water content of the liquid produced by the wellhead of the high-yield gas-oil well based on the multi-sensor measurement and the deep convolutional neural network as claimed in claim 1, wherein the method comprises the following steps: in the step (3), the deep convolutional neural network comprises 5 convolutional layers, wherein pooling is performed on the first convolutional layer, the second convolutional layer and the fifth convolutional layer to prevent an over-fitting phenomenon; the first layer of convolution operation adopts 48 convolution kernels with the size of 11 x 11, the step size is set to be 4, and then pooling operation is carried out, the size of the pooling convolution kernel is 3 x 3, and the step size is 2; the second layer of convolution adopts 128 convolution kernels with the size of 5 x 5, the step size is set to be 1, the size of the pooling convolution kernel is 3 x 3, and the step size is 2; then, the third layer and the fourth layer are convolution layers, the pooling operation is not carried out, the sizes of convolution kernels are all set to be 3 x 3, and the number of convolution kernels is all set to be 192; the fifth layer is a convolution layer, 128 convolution kernels with the size of 3 x 3 are arranged, the step size is 1, the size of the pooling convolution kernel is 3 x 3, and the step size is 2; features extracted by the deep convolutional neural network are input into a two-layer full-connection network, the number of nodes of each layer of the full-connection network is 1024, and finally, the water content high-abstraction features are input into a softmax classifier to predict the water content.
5. The method for predicting the water content of the liquid produced by the wellhead of the high-yield gas-oil well based on the multi-sensor measurement and the deep convolutional neural network as claimed in claim 1, wherein the method comprises the following steps: the structure of the sensor assembly is as follows: the two ends of the stainless steel protective shell of the sensor are provided with a left flange and a right flange, wherein the right flange is connected with the stainless steel protective shell in a threaded engagement mode; left flange and right flange and well head descent pipe connection, the inside coaxial arrangement of stainless steel protective housing has the woollen goods pipeline, interval installation has two annular sensor measuring electrode on woollen goods pipeline outer wall, coaxial arrangement has the electromagnetic shield layer in the sensor measuring electrode outside, the woollen goods pipeline compresses tightly sealedly with the stainless steel protective housing through the O type circle on top, still install transmission-type ultrasonic sensor on woollen goods pipeline outer wall, open the stainless steel protective housing lateral wall has the pin hole, organic glass ring support between electromagnetic shield layer and the woollen goods pipeline.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910613946.8A CN110632132A (en) | 2019-07-09 | 2019-07-09 | High-yield gas-oil well wellhead liquid water content prediction method based on multi-sensor measurement and deep convolutional neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910613946.8A CN110632132A (en) | 2019-07-09 | 2019-07-09 | High-yield gas-oil well wellhead liquid water content prediction method based on multi-sensor measurement and deep convolutional neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110632132A true CN110632132A (en) | 2019-12-31 |
Family
ID=68969242
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910613946.8A Pending CN110632132A (en) | 2019-07-09 | 2019-07-09 | High-yield gas-oil well wellhead liquid water content prediction method based on multi-sensor measurement and deep convolutional neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110632132A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111310900A (en) * | 2020-02-24 | 2020-06-19 | 姜景元 | Multiphase flow detection method, device, equipment and medium |
| CN111638249A (en) * | 2020-05-31 | 2020-09-08 | 天津大学 | Water content measuring method based on deep learning and application of water content measuring method in oil well exploitation |
| CN112814649A (en) * | 2021-01-04 | 2021-05-18 | 天津科技大学 | Intelligent metering device and method for water content of oil well produced liquid |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE2211997A1 (en) * | 1971-03-17 | 1972-09-21 | United Aircraft Corp | Method for separating a mixture of two liquids with different densities and a device for carrying out the method |
| CN1601265A (en) * | 2004-10-26 | 2005-03-30 | 大庆油田有限责任公司 | Analyzer for analyzing moisture in ground conductance |
| CN102587898A (en) * | 2012-03-08 | 2012-07-18 | 中国石油天然气集团公司 | Method and device for detecting gas content of mixed fluid under while drilling condition |
| CN108229659A (en) * | 2017-12-29 | 2018-06-29 | 陕西科技大学 | Piano singly-bound voice recognition method based on deep learning |
| CN108805039A (en) * | 2018-04-17 | 2018-11-13 | 哈尔滨工程大学 | The Modulation Identification method of combination entropy and pre-training CNN extraction time-frequency image features |
| CN109343060A (en) * | 2018-12-07 | 2019-02-15 | 电子科技大学 | ISAR imaging method and system based on deep learning time frequency analysis |
-
2019
- 2019-07-09 CN CN201910613946.8A patent/CN110632132A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE2211997A1 (en) * | 1971-03-17 | 1972-09-21 | United Aircraft Corp | Method for separating a mixture of two liquids with different densities and a device for carrying out the method |
| CN1601265A (en) * | 2004-10-26 | 2005-03-30 | 大庆油田有限责任公司 | Analyzer for analyzing moisture in ground conductance |
| CN102587898A (en) * | 2012-03-08 | 2012-07-18 | 中国石油天然气集团公司 | Method and device for detecting gas content of mixed fluid under while drilling condition |
| CN108229659A (en) * | 2017-12-29 | 2018-06-29 | 陕西科技大学 | Piano singly-bound voice recognition method based on deep learning |
| CN108805039A (en) * | 2018-04-17 | 2018-11-13 | 哈尔滨工程大学 | The Modulation Identification method of combination entropy and pre-training CNN extraction time-frequency image features |
| CN109343060A (en) * | 2018-12-07 | 2019-02-15 | 电子科技大学 | ISAR imaging method and system based on deep learning time frequency analysis |
Non-Patent Citations (4)
| Title |
|---|
| HLAWATSCH F 等: "Smoothed pseudo-wigner distribution,choi-williams distribution and cone-kernel representation:ambiguity-domain analysis and experimental comparison", 《SIGNAL PROCESSING》 * |
| 张宁: "自适应时频分析及其时频属性提取方法研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 汪源源: "《现代信号处理理论和方法》", 2003083 * |
| 王丽娜 等: "基于CPN网络井口计量原油含水率预测模型", 《大庆石油学院学报》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111310900A (en) * | 2020-02-24 | 2020-06-19 | 姜景元 | Multiphase flow detection method, device, equipment and medium |
| CN111638249A (en) * | 2020-05-31 | 2020-09-08 | 天津大学 | Water content measuring method based on deep learning and application of water content measuring method in oil well exploitation |
| CN111638249B (en) * | 2020-05-31 | 2022-05-17 | 天津大学 | Water content measuring method based on deep learning and application of water content measuring method in oil well exploitation |
| CN112814649A (en) * | 2021-01-04 | 2021-05-18 | 天津科技大学 | Intelligent metering device and method for water content of oil well produced liquid |
| CN112814649B (en) * | 2021-01-04 | 2023-03-10 | 天津科技大学 | Intelligent metering device and method for water content of oil well produced liquid |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Miao et al. | Online defect recognition of narrow overlap weld based on two-stage recognition model combining continuous wavelet transform and convolutional neural network | |
| CN110632132A (en) | High-yield gas-oil well wellhead liquid water content prediction method based on multi-sensor measurement and deep convolutional neural network | |
| CN110630256A (en) | Low-gas-production oil well wellhead water content prediction system and method based on depth time memory network | |
| CN104101687B (en) | Based on the polyphasic flow visual testing method of ultrasonic Doppler and electricity multisensor | |
| CN101609113B (en) | Non-contact conductivity measuring device based on double shield structure and acceptor resonance and method thereof | |
| US20140305194A1 (en) | System for measurement of fluid levels in multi-phase fluids | |
| CN104662415A (en) | Systems and methods for measuring an interface level in a multiphase fluid composition | |
| 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 | |
| CN106461449A (en) | Sensor systems for measuring an interface level in a multi-phase fluid composition | |
| CN109060771A (en) | A kind of common recognition model building method based on spectrum different characteristic collection | |
| CN106525755A (en) | Oil-sand pH value testing method based on near infrared spectroscopy technology | |
| CN110514366B (en) | Method for detecting weak leakage of pipeline under small sample condition | |
| CN113376609B (en) | Liquid identification method based on millimeter wave radar | |
| CN110243885A (en) | A kind of low yield gas well mouth of oil well hydrated comples ion method based on time-frequency characteristics | |
| CN112814649B (en) | Intelligent metering device and method for water content of oil well produced liquid | |
| Chen et al. | Water cut measurement of oil–water flow in vertical well by combining total flow rate and the response of a conductance probe | |
| CN110243886A (en) | A kind of low yield gas well mouth of oil well hydrated comples ion method based on nonlinear characteristic | |
| CN116337946A (en) | Device and method for measuring water content of crude oil based on GRU neural network model | |
| Zhao et al. | Research on Transformer Oil Multi-frequency Ultrasonic Monitoring Technology Based on Convolutional Neural Network | |
| CN115407412A (en) | Time domain electromagnetic wave detection method for well-side abnormal body | |
| CN110231410B (en) | Intelligent interpretation method for nondestructive testing data of anchor rod | |
| Ding et al. | Recognition method of coal–rock reflection spectrum using wavelet scattering transform and bidirectional long–short-term memory | |
| CN201421479Y (en) | Non-contact conductance measuring device based on double shielding structure and series resonance | |
| Ying et al. | An improved feature parameter extraction algorithm of composite detection method based on the fusion theory | |
| Feng et al. | Automatic deep vector learning model applied for oil-well-testing feature mining, purification and classification |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191231 |
|
| RJ01 | Rejection of invention patent application after publication |