Wellhead three-phase content metering method based on multi-parameter fusion

Abstract

The application discloses a kind of wellhead three-phase content rate measurement methods based on multi-parameter fusion, temperature and pressure electric multi-parameter fusion network includes: temperature and pressure NN-LSTM module, microwave Transform module, attention multi-parameter aggregation module and prediction output module, the input sample of temperature and pressure electric multi-parameter fusion network is temperature related signal, pressure related signal, microwave difference frequency phase related signal and microwave difference frequency amplitude related signal, and output is the water content rate prediction value of sample.The wellhead three-phase content rate measurement method of the application: obtain the sample to be predicted;Using the temperature and pressure electric multi-parameter fusion network trained, the water content rate prediction value of the sample to be predicted is obtained by predicting the sample to be predicted.The wellhead three-phase content rate measurement method of the application accurately predicts the water content rate of oil wellhead produced liquid.

Description

Technical Field

[0001] This invention belongs to the field of oil well parameter measurement, specifically relating to a wellhead three-phase content measurement method based on multi-parameter fusion. Background Technology

[0002] In oilfield development, accurately measuring the water cut of the three-phase fluid at the wellhead is crucial for production scheduling, equipment maintenance, and optimized management. Currently, traditional manual sampling and testing methods are labor-intensive, and while commonly used sensors based on conductivity, capacitance, and radiation sources offer some monitoring accuracy, they are costly. Furthermore, measurement methods relying solely on artificial intelligence often suffer from insufficient measurement accuracy in practical applications.

[0003] With the continuous development of sensor technology, the application of microwave sensors in wellhead fluid monitoring has been gradually promoted. By measuring the amplitude and phase changes of microwave signals, the composition of fluids can be indirectly reflected. However, single sensors often cannot meet the requirements of high precision, and for complex changes in wellhead fluid state, traditional data processing methods based on single sensors cannot fully tap the potential of multi-source information.

[0004] Therefore, there is an urgent need for a high-precision, real-time wellhead three-phase content measurement method. Summary of the Invention

[0005] To address the shortcomings of existing technologies, the purpose of this invention is to provide a thermo-piezoelectric multi-parameter fusion network.

[0006] Another objective of this invention is to provide a wellhead three-phase water cut measurement method based on multi-parameter fusion. This wellhead three-phase water cut measurement method enables a thermo-piezoelectric multi-parameter fusion network to accurately predict the water cut of oil wellhead produced fluids, thereby achieving high-precision prediction and real-time monitoring of the water cut of oil wellhead produced fluids.

[0007] This invention is achieved through the following technical solution.

[0008] A thermo-piezoelectric multi-parameter fusion network includes: a thermo-piezoelectric CNN-LSTM module, a microwave Transformer module, an attention multi-parameter aggregation module, and a prediction output module;

[0009] The temperature-pressure CNN-LSTM module includes: a first CNN-LSTM module and a second CNN-LSTM module. The input signal of the first CNN-LSTM module is a temperature-related signal, and the output is the characteristics of the temperature-related signal. The input signal of the second CNN-LSTM module is a pressure-related signal, and the output is the characteristics of the pressure-related signal.

[0010] The microwave Transformer module includes: a first Transformer model and a second Transformer model. The first Transformer model takes microwave difference frequency phase correlation signal as input and outputs the characteristics of microwave difference frequency phase correlation signal as output. The second Transformer model takes microwave difference frequency amplitude correlation signal as input and outputs the characteristics of microwave difference frequency amplitude correlation signal as output.

[0011] The attention multi-parameter aggregation module uses an attention fusion mechanism to weight and fuse the features of temperature-related signals, pressure-related signals, microwave difference frequency phase-related signals, and microwave difference frequency amplitude-related signals to obtain fused features.

[0012] The temperature-related signal, pressure-related signal, microwave difference frequency phase-related signal, and microwave difference frequency amplitude-related signal are, in order, the standardized signals of the temperature signal, pressure signal, microwave difference frequency phase signal, and microwave difference frequency amplitude signal. The temperature signal, pressure signal, microwave difference frequency phase signal, and microwave difference frequency amplitude signal are obtained by the corresponding sensors from the oil wellhead produced fluid in the same time period.

[0013] The input to the prediction output module is the fused features, and the output value is the predicted water cut of the oil wellhead produced fluid collected during that time period. The prediction output module uses Sigmoid as the activation function.

[0014] A wellhead three-phase fill rate measurement method based on multi-parameter fusion includes:

[0015] Step 1: Obtain the temperature correlation signal, pressure correlation signal, microwave difference frequency phase correlation signal, and microwave difference frequency amplitude correlation signal of the sample to be predicted;

[0016] Step 2: Use the trained thermo-piezoelectric multi-parameter fusion network to predict the sample to be predicted and obtain the predicted water content value of the sample.

[0017] In the above technical solution, the method for obtaining the trained temperature-piezoelectric multi-parameter fusion network includes: preparing multiple samples as a training set, each sample including: temperature-related signal, pressure-related signal, microwave difference frequency phase-related signal, and microwave difference frequency amplitude-related signal; substituting the samples into the temperature-piezoelectric multi-parameter fusion network for training; substituting the predicted moisture content value of the sample and the moisture content label value of the sample into the loss function to obtain the loss function value; calculating the gradient using the AMSGrad optimization algorithm based on the loss function value; and updating the weights of the temperature-piezoelectric multi-parameter fusion network through the gradient backpropagation method.

[0018] In the above technical solution, the water content label value is the true value of the water content of the oil wellhead produced fluid during the sample collection period.

[0019] In the above technical solution, the loss function is the mean squared error.

[0020] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0021] 1. The wellhead three-phase water cut measurement method of the present invention integrates signals from multiple sensors, overcomes the limitations of a single sensor, improves the accuracy and reliability of prediction, accurately predicts the water cut of oil wellhead produced fluid, and has high prediction accuracy.

[0022] 2. The wellhead three-phase water cut measurement method of the present invention realizes the prediction of the water cut of the produced fluid at the oil wellhead through a multi-parameter fusion network of temperature, piezoelectricity, and electromechanical parameters, which effectively improves the accuracy and real-time performance of wellhead fluid monitoring and is of great significance to oilfield production management and equipment maintenance. Attached Figure Description

[0023] Figure 1 This is a structural diagram of a thermo-piezoelectric multi-parameter fusion network;

[0024] Figure 2 This is a structural diagram of a multi-sensor measurement system;

[0025] Figure 3 This is a schematic diagram of a multi-sensor measurement system.

[0026] 1: Oil inlet pipe section; 2: Opposite microwave sensor; 3: Temperature sensor; 4: Pressure sensor; 5: Oil outlet pipe section; 6: Casing. Detailed Implementation

[0027] The wellhead three-phase content measurement method of the present invention will be described in detail below with reference to the accompanying drawings.

[0028] Example 1

[0029] A thermo-piezoelectric multi-parameter fusion network, such as Figure 1 As shown, it includes: a thermo-pressure CNN-LSTM module, a microwave Transformer module, an attention multi-parameter aggregation module, and a prediction output module;

[0030] The temperature-pressure CNN-LSTM module includes: a first CNN-LSTM module and a second CNN-LSTM module (see: Li Mei, Ning Dejun, Guo Jiacheng. CNN-LSTM model based on attention mechanism and its application [J]. Computer Engineering and Applications, 2019, 55(13):20-27.). The input signal of the first CNN-LSTM module is a temperature-related signal and the output is the features of the temperature-related signal. The input signal of the second CNN-LSTM module is a pressure-related signal and the output is the features of the pressure-related signal. Both the first CNN-LSTM module and the second CNN-LSTM module are used to extract the local spatial features and temporal features of the input signal.

[0031] Specifically, the first CNN-LSTM module uses a CNN (Convolutional Neural Network) to extract local features from temperature-related signals, capturing local patterns and regularities to obtain feature X1. Then, a ReLU activation function is used to perform a nonlinear transformation on feature X1 to obtain feature X2. Finally, an LSTM (Long Short-Term Memory) network is used to capture the temporal relationships in feature X2. The LSTM updates the hidden state through a gating mechanism to capture the changes in temperature-related signals over time, thus obtaining the features of the temperature-related signals. The second CNN-LSTM module uses a CNN to extract local features from pressure-related signals, capturing local patterns and regularities to obtain feature X3. Then, a ReLU activation function is used to perform a nonlinear transformation on feature X3 to obtain feature X4. Finally, an LSTM is used to capture the temporal relationships in feature X4. The LSTM updates the hidden state through a gating mechanism to capture the changes in pressure-related signals over time, thus obtaining the features of the pressure-related signals.

[0032] The microwave Transformer module includes: a first Transformer model and a second Transformer model (see: Fu Yan, Yang Xu, Ye Ou. Smoke Recognition Method Based on CNN and Transformer Feature Fusion [J]. Computer Engineering and Science, 2024, 46(11): 2045-2052.). The input signal of the first Transformer model is the microwave difference frequency phase correlation signal, and the output is the feature of the microwave difference frequency phase correlation signal; the input signal of the second Transformer model is the microwave difference frequency amplitude correlation signal, and the output is the feature of the microwave difference frequency amplitude correlation signal; both the first Transformer model and the second Transformer model are used to perform sequence modeling and feature extraction on the input signal.

[0033] The attention multi-parameter aggregation module uses an attention fusion mechanism (Wang Xiaolan, Zhang Weidong, Wang Huizhong. Short-term load prediction based on attention mechanism of CNN-LSTM[J]. Computer and Digital Engineering, 2024, 52(10):3014-3018.) to weight and fuse the features of temperature-related signals and pressure-related signals output by the temperature and pressure CNN-LSTM module, as well as the features of microwave difference frequency phase-related signals and microwave difference frequency amplitude-related signals output by the microwave Transformer module, to obtain fused features;

[0034] The temperature-related signal, pressure-related signal, microwave difference frequency phase-related signal, and microwave difference frequency amplitude-related signal are, in order, the standardized signals of the temperature signal, pressure signal, microwave difference frequency phase signal, and microwave difference frequency amplitude signal. The temperature signal, pressure signal, microwave difference frequency phase signal, and microwave difference frequency amplitude signal are obtained by the corresponding sensors from the oil wellhead produced fluid in the same time period.

[0035] The input to the prediction output module is the fused features, and the output value is the predicted water cut of the oil wellhead produced fluid collected during that time period. The prediction output module uses Sigmoid as the activation function.

[0036] Example 2

[0037] Based on Example 1, a multi-sensor measurement system is used to acquire temperature signals, pressure signals, microwave difference frequency phase signals, and microwave difference frequency amplitude signals, such as... Figure 2 and Figure 3 As shown, the multi-sensor measurement system includes: an inlet pipe section 1, a vertical microwave sensor 2, a temperature sensor 3, a pressure sensor 4, an outlet pipe section 5, and a casing 6. One end of the inlet pipe section 1 is connected to the wellhead, and the casing 6 is fitted over one end of the inlet pipe section 1. The vertical microwave sensor 2, temperature sensor 3, and pressure sensor 4 are installed on the inlet pipe section 1 and located on one side outside the casing 6. One end of the outlet pipe section 5 is connected to the casing, and the other end of the outlet pipe section 5 is connected to the inlet pipe section 1. The wellhead produced fluid (three-phase fluid) enters the multi-sensor measurement system through the inlet pipe section 1; the vertical microwave sensor collects the microwave difference frequency phase signal and microwave difference frequency amplitude signal of the wellhead produced fluid; the temperature sensor collects the temperature signal of the wellhead produced fluid; and the pressure sensor collects the pressure signal of the wellhead produced fluid. The wellhead produced fluid flows into the casing 6 through the outlet pipe section and is then discharged.

[0038] Preparing a multi-parameter dataset involves the following steps:

[0039] S1, the multi-sensor measurement system collects sample data once every T1 interval, for a total of N collections. The time for each collection is T2. The sample data collected each time includes: microwave difference frequency phase signal, microwave difference frequency amplitude signal, temperature signal and pressure signal. During the T2 time of each sample data collection by the multi-sensor measurement system, M manual samples of oil wellhead produced fluid (three-phase fluid) are taken at time intervals of T3, for a total of M×N manual samples. All manually sampled oil wellhead produced fluid (three-phase fluid) are manually tested to detect the water content of the oil wellhead produced fluid, and the results are used as test values ​​to obtain M×N test values.

[0040] In the multi-sensor measurement system, the acquisition start time of the opposing microwave sensor, temperature sensor and pressure sensor are all the same, and the sampling frequency is F=1000Hz; T1=1h, T2=300s, T3=60s, N=100, M=5, M×N=500, and the number of test values ​​is 500.

[0041] S2, the microwave difference frequency phase signal, microwave difference frequency amplitude signal, temperature signal and pressure signal acquired by the multi-sensor measurement system are used as acquisition signals, and each acquisition signal is preprocessed:

[0042] S2-1, using deviation standardization (see: Xu Yifang, Chen Jin, Li Lin, et al. Credit prediction based on Isomap fusion Naive Bayes classifier [J]. Computer Knowledge and Technology, 2021, 17(35): 125-126+139.) to scale the collected signal to the range of [0, 1] to obtain the standardized signal;

[0043] S2-2, using the sliding window method (Zhao Yuxuan, Gong Jianning, Liu Han, et al. Research on front-end processing of mobile robots based on multi-source sensor information fusion [J]. Manufacturing Automation, 2023, 45(04): 217-220.), the standardized signal is segmented according to the acquisition time sequence, and the sliding window length is set to L = 6 × 10 4 And without overlapping segmentation, each standardized signal is divided into segments according to the acquisition time sequence. A related signal.

[0044] When the acquired signal is a microwave difference frequency phase signal, the correlated signal is a microwave difference frequency phase correlation signal;

[0045] When the acquired signal is a microwave difference frequency amplitude signal, the correlated signal is a microwave difference frequency amplitude correlated signal;

[0046] When the acquired signal is a temperature signal, the correlated signal is a temperature-related signal;

[0047] When the acquired signal is a pressure signal, the related signal is a pressure-related signal.

[0048] Temperature-related signals, pressure-related signals, microwave difference-frequency phase-related signals, and microwave difference-frequency amplitude-related signals obtained within the same time period are considered as one sample. All samples constitute a multi-parameter dataset; that is, in this embodiment, the multi-parameter dataset contains a total of [number missing]. One sample.

[0049] The M×N test values ​​are arranged in the order of manual sampling time as... The water content label value of the oil wellhead produced fluid corresponding to each sample.

[0050] Example 3

[0051] Based on Example 2, a training method for a thermo-piezoelectric multi-parameter fusion network includes: training and validating the thermo-piezoelectric multi-parameter fusion network using a 10-fold cross-validation method on the multi-parameter dataset to obtain a trained thermo-piezoelectric multi-parameter fusion network. Nine samples are used as the training set, and one sample is used as the validation set, so that all samples in the multi-parameter dataset are used as the validation set.

[0052] Each training iteration consists of 100 rounds. During each training iteration, the thermo-piezoelectric multi-parameter fusion network predicts the moisture content of the training set and outputs the predicted moisture content value. The loss function is calculated based on the predicted moisture content value and the moisture content label value. The gradient is then calculated using the AMSGrad optimization algorithm (Li Manyuan, Luo Fei, Gu Chunhua, et al. Adams algorithm based on adaptive momentum update strategy [J]. Journal of Shanghai University of Science and Technology, 2023, 45(02): 112-119.) based on the loss function value. The weights of the thermo-piezoelectric multi-parameter fusion network are updated through the gradient backpropagation method.

[0053] The loss function is the mean squared error (MSE), which measures the difference between the predicted moisture content and the labeled moisture content output by the thermo-piezoelectric multi-parameter fusion network. The gradient backpropagation method uses minimizing the loss function as a criterion to guide the training and optimization of the thermo-piezoelectric multi-parameter fusion network in the right direction.

[0054] Example 4

[0055] A wellhead three-phase fill rate measurement method based on multi-parameter fusion includes:

[0056] Step 1: Obtain temperature correlation signals, pressure correlation signals, microwave difference frequency phase correlation signals, microwave difference frequency amplitude correlation signals, and moisture content label values ​​of 50 samples to be predicted;

[0057] Step 2: Using the thermo-piezoelectric multi-parameter fusion network trained in Example 3, predict the moisture content of the sample to be predicted.

[0058] Example 5

[0059] A wellhead three-phase content measurement method based on multi-parameter fusion is basically the same as that in Example 4, except that the temperature-pressure CNN-LSTM module in the temperature-pressure-electric multi-parameter fusion network used in Example 4 is removed in this example.

[0060] Example 6

[0061] A wellhead three-phase content measurement method based on multi-parameter fusion is basically the same as that in Example 4, except that the microwave Transformer module in the thermo-piezoelectric multi-parameter fusion network used in Example 4 is removed in this example.

[0062] The prediction results of Examples 4, 5, and 6 were evaluated using mean squared error (MSE), mean absolute error (MAE), and mean absolute percentage error (MAPE) as follows:

[0063] Table 1

[0064] Model MSE MAE MAPE Example 5 0.00428 0.0637 0.1086 Example 6 0.0334 0.437 0.856 Example 4 0.00206 0.0389 0.0593

[0065] As shown in Table 1, the water cut predicted by the wellhead three-phase water cut metering method in Example 4 has a low error and high prediction accuracy.

[0066] The present invention has been described above by way of example. It should be noted that any simple modifications, alterations or other equivalent substitutions that can be made by those skilled in the art without creative effort without departing from the core of the present invention fall within the protection scope of the present invention.

Claims

1. A warm pressure electric multi-parameter fusion network, characterized in that, The method comprises the following steps: The temperature-pressure CNN-LSTM module comprises a first CNN-LSTM module and a second CNN-LSTM module, wherein the input signal of the first CNN-LSTM module is a temperature-related signal, and the output is a feature of the temperature-related signal; the input signal of the second CNN-LSTM module is a pressure-related signal, and the output is a feature of the pressure-related signal; The microwave Transformer module comprises a first Transformer model and a second Transformer model, wherein the input signal of the first Transformer model is a microwave difference frequency phase-related signal, and the output is a feature of the microwave difference frequency phase-related signal; the input signal of the second Transformer model is a microwave difference frequency amplitude-related signal, and the output is a feature of the microwave difference frequency amplitude-related signal; The attention multi-parameter aggregation module adopts an attention fusion mechanism to perform weighted fusion on the features of the temperature-related signal and the pressure-related signal, the features of the microwave difference frequency phase-related signal and the microwave difference frequency amplitude-related signal, and obtain a fused feature; The temperature-related signal, the pressure-related signal, the microwave difference frequency phase-related signal and the microwave difference frequency amplitude-related signal are, in sequence, a signal standardized from a temperature signal, a signal standardized from a pressure signal, a signal standardized from a microwave difference frequency phase signal and a signal standardized from a microwave difference frequency amplitude signal, wherein the temperature signal, the pressure signal, the microwave difference frequency phase signal and the microwave difference frequency amplitude signal are obtained by corresponding sensors from oil wellhead production fluid collected in the same time period; The input of the prediction output module is the fused feature, and the output value is used as a water cut prediction value of the oil wellhead production fluid collected in the time period; the prediction output module uses Sigmoid as an activation function. The method comprises the following steps:

2. A wellhead three-phase inclusion rate measurement method based on multi-parameter fusion, characterized in that, Step 1: obtaining a temperature-related signal, a pressure-related signal, a microwave difference frequency phase-related signal and a microwave difference frequency amplitude-related signal of a sample to be predicted; Step 2: using the temperature-pressure electric multi-parameter fusion network in claim 1 trained to predict the sample to be predicted, and obtaining a water cut prediction value of the sample to be predicted. The method for obtaining the trained temperature-pressure electric multi-parameter fusion network comprises the following steps:

3. The method for measuring the wellhead three-phase content rate based on multi-parameter fusion according to claim 2, characterized in that, A plurality of samples are prepared as a training set, each sample comprising a temperature-related signal, a pressure-related signal, a microwave difference frequency phase-related signal and a microwave difference frequency amplitude-related signal; the samples are input into the temperature-pressure electric multi-parameter fusion network for training; a loss function value is obtained by inputting a water cut prediction value of the sample and a water cut label value of the sample into a loss function; a gradient is calculated by using an AMSGrad optimization algorithm according to the loss function value; and the weights of the temperature-pressure electric multi-parameter fusion network are updated by using a gradient back propagation method. The water cut label value is a true value of the water cut of the oil wellhead production fluid collected in the sample time period.

4. The method for measuring the wellhead three-phase content rate based on multi-parameter fusion according to claim 3, characterized in that, The loss function is a mean square error function.

5. The method for measuring the wellhead three-phase content rate based on multi-parameter fusion according to claim 3, characterized in that, ​

Images