A method, device, equipment, storage medium and product for predicting load of a pumping well

Abstract

The application discloses a pumping unit well load prediction method, device, equipment, storage medium and product, relates to the technical field of industrial automation and energy efficiency optimization, and the method comprises the steps of obtaining static parameters and load parameters of the pumping unit well at different times under the same working condition, obtaining a mapping relationship between the static parameters and the load parameters through a first prediction model, and the mapping relationship representing discrete characteristics of pumping unit well production; based on the mapping relationship and the static parameters of the pumping unit well at different times under the same working condition, the first prediction model is used to predict the pumping unit load parameters at corresponding times under the same working condition, and a first load prediction curve is obtained; a historical load curve under the same working condition is input into a second prediction model, and a second load prediction curve is obtained, and the historical load curve represents continuous production characteristics of the pumping unit well; and according to the first load prediction curve and the second load prediction curve, a target prediction curve of the pumping unit well is obtained. The combination of the two prediction curves can reduce errors and improve prediction accuracy.

Description

Technical Field

[0001] This application relates to the fields of industrial automation and energy efficiency optimization technology, and in particular to load prediction methods, devices, equipment, storage media and products for oil pumping wells. Background Technology

[0002] Pumping units are the most important energy-consuming equipment in oilfield development, with pumping wells accounting for over 90% of all pumped wells. Pumping systems account for more than 70% of energy consumption in oil and gas production. Due to the randomness, volatility, and unpredictability of renewable energy generation, the proportion of clean energy directly replacing renewable energy in oil and gas production is affected. Accurate load forecasting helps companies more accurately predict future energy demand, thereby better planning and allocating resources and reducing energy waste.

[0003] Current IoT standards for oil and gas production do not require continuous data collection of pumping unit well loads. Instead, they record instantaneous values ​​at regular intervals. However, discrete power values ​​cannot accurately reflect the load status, characteristics, and even production status of pumping unit wells, resulting in low accuracy in predicting pumping unit loads.

[0004] Therefore, improving the accuracy of oil pumping unit load prediction is an urgent problem to be solved. Summary of the Invention

[0005] The main objective of this application is to provide a method, apparatus, equipment, storage medium, and product for predicting the load of pumping wells, aiming to solve the technical problem of low prediction accuracy of pumping well load.

[0006] To achieve the above objectives, this application proposes a method for predicting the load of pumping wells, the method comprising:

[0007] The static parameters and load parameters of the pumping unit well at different times under the same operating conditions are obtained. The mapping relationship between the static parameters and the load parameters is obtained through the first prediction model. The mapping relationship is used to represent the discrete production characteristics of the pumping unit well.

[0008] Based on the mapping relationship and the static parameters of the pumping unit well at different times under the same working condition, the pumping unit load parameters at the corresponding time under the same working condition are predicted by the first prediction model to obtain the first load prediction curve.

[0009] The historical load curve under the same operating condition is input into the second prediction model to obtain the second load prediction curve, which is used to represent the continuous production characteristics of the pumping well.

[0010] Based on the first load prediction curve and the second load prediction curve, the target prediction curve for the pumping well is obtained.

[0011] In one embodiment, the step of obtaining the static parameters and load parameters of the pumping unit well at different times under the same operating conditions, and obtaining the mapping relationship between the static parameters and the load parameters, includes:

[0012] The static parameters and load parameters of the pumping well under the same working condition at different times are preprocessed, and the preprocessed data is input into the input layer of the first prediction model to obtain the first dataset;

[0013] The first dataset is input into the hidden layer, and the mapping relationship between static parameters and load parameters is extracted through a multi-layer neural network structure;

[0014] The mapping relationship between the static parameters and the load parameters is output through the output layer.

[0015] In one embodiment, the step of predicting the pumping unit load parameters at corresponding times under the same operating condition using a first prediction model based on the mapping relationship and the static parameters of the pumping unit well at different times under the same operating condition, and obtaining a first load prediction curve, includes:

[0016] Based on the mapping relationship and the static parameters of the pumping unit well at different times under the same working condition, feature information corresponding to the static parameters of the pumping unit well at different times under the same working condition is extracted through the hidden layer.

[0017] The feature information is nonlinearly mapped by an activation function to generate predicted load parameters corresponding to the feature information.

[0018] Based on the predicted load parameters, the first load prediction curve under the same operating condition is obtained.

[0019] In one embodiment, the step of inputting historical load curves under the same operating conditions into a second prediction model to obtain a second load prediction curve includes:

[0020] The historical load curves under the same operating conditions are input into the input layer of the second prediction model to obtain the second dataset;

[0021] The second dataset is input into the hidden layer of the second prediction model to generate load forecast values ​​corresponding to the historical time series;

[0022] The load forecast values ​​at multiple time steps are reconstructed to obtain a second load forecast curve.

[0023] In one embodiment, the step of inputting the second dataset into the hidden layer of the second prediction model to generate load forecast values ​​corresponding to historical time series includes:

[0024] The second dataset is input into the long short-term memory unit, and historical load information is extracted through the memory gate, input gate and forget gate mechanisms;

[0025] The historical load information is input into the hidden layer of the second prediction model, and time-dependent features are extracted through multi-layer long short-term memory units. The time-dependent features capture the continuous production characteristics of the pumping well.

[0026] The time-dependent features are input into a fully connected layer to generate load forecast values ​​corresponding to historical time series.

[0027] In one embodiment, the step of obtaining the target prediction curve of the pumping well based on the first load prediction curve and the second load prediction curve includes:

[0028] The target predicted value of the pumping well is obtained by averaging and superimposing the predicted values ​​at corresponding time points of the first load prediction curve and the second load prediction curve.

[0029] Based on the target predicted value of the pumping well, the target prediction curve of the pumping well is obtained.

[0030] Furthermore, to achieve the above objectives, this application also proposes a load prediction device for a pumping unit well, the load prediction device for the pumping unit well comprising:

[0031] The mapping relationship acquisition module is used to acquire the static parameters and load parameters of the pumping well at different times under the same working condition, and obtain the mapping relationship between the static parameters and load parameters through the first prediction model. The mapping relationship is used to represent the discrete production characteristics of the pumping well.

[0032] The first prediction curve module is used to predict the pumping unit load parameters at the corresponding time under the same working condition based on the mapping relationship and the static parameters of the pumping unit well at different times under the same working condition, and obtain the first load prediction curve.

[0033] The second prediction curve module is used to input the historical load curve under the same operating condition into the second prediction model to obtain the second load prediction curve. The historical load curve is used to represent the continuous production characteristics of the pumping well.

[0034] The target prediction curve module is used to obtain the target prediction curve of the pumping well based on the first load prediction curve and the second load prediction curve.

[0035] In addition, to achieve the above objectives, this application also proposes a load prediction device for pumping wells, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the load prediction method for pumping wells as described above.

[0036] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the load prediction method for pumping wells as described above.

[0037] In addition, to achieve the above objectives, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the load prediction method for pumping wells as described above.

[0038] One or more technical solutions proposed in this application have at least the following technical effects:

[0039] The process involves acquiring static and load parameters of a pumping unit well under the same operating conditions at different times. A first prediction model is used to obtain the mapping relationship between these static and load parameters, representing the discrete production characteristics of the pumping unit well. Based on this mapping relationship and the static parameters of the pumping unit well under the same operating conditions at different times, the first prediction model predicts the pumping unit load parameters at the corresponding time under the same operating conditions, resulting in a first load prediction curve. Historical load curves under the same operating conditions are input into a second prediction model to obtain a second load prediction curve, representing the continuous production characteristics of the pumping unit well. Finally, based on the first and second load prediction curves, a target prediction curve for the pumping unit well is obtained. The first load prediction curve reflects the changes in the production state of the pumping unit well at a specific time, while the second load prediction curve, based on the historical load curve, derives the long-term trend using continuous production characteristics. The combination of these two prediction curves effectively reduces the error caused by data limitations in a single prediction model, improving the accuracy and reliability of the prediction. Attached Figure Description

[0040] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0041] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0042] Figure 1 This is a flowchart illustrating the first embodiment of the load prediction method for pumping wells in this application.

[0043] Figure 2 This is a flowchart illustrating the second embodiment of the load prediction method for pumping wells in this application.

[0044] Figure 3 This is a flowchart illustrating the third embodiment of the load prediction method for pumping wells in this application.

[0045] Figure 4 This is a flowchart illustrating the third embodiment of the load prediction method for pumping wells in this application.

[0046] Figure 5 This is a schematic diagram of the original load curve of an embodiment of this application;

[0047] Figure 6 This is a schematic diagram of the historical load curve before the smoothing operation in an embodiment of this application;

[0048] Figure 7 This is a schematic diagram of the historical load curve after the smoothing operation in an embodiment of this application;

[0049] Figure 8 This is a schematic diagram of the second load prediction curve according to an embodiment of this application;

[0050] Figure 9 This is a schematic diagram of the module structure of the load prediction device for a pumping well according to an embodiment of this application;

[0051] Figure 10 This is a schematic diagram of the equipment structure of the hardware operating environment involved in the load prediction method for pumping wells in this application embodiment.

[0052] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0053] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0054] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0055] Due to the randomness, volatility, and unpredictability of new energy power generation, the proportion of clean energy directly replacing oil and gas production is affected. Accurate load forecasting can help companies predict future energy demand more precisely. Current IoT standards for oil and gas production do not require continuous data collection of pumping unit well loads; they only record instantaneous values ​​at regular intervals. However, discrete power values ​​cannot accurately reflect the load status, characteristics, and even production status of pumping unit wells, resulting in low accuracy in pumping unit load forecasting.

[0056] This application provides a solution to obtain the static and load parameters of a pumping unit well under the same operating conditions at different times. A first prediction model is used to obtain the mapping relationship between the static and load parameters, representing the discrete production characteristics of the pumping unit well. Based on the mapping relationship and the static parameters of the pumping unit well under the same operating conditions at different times, the first prediction model predicts the pumping unit load parameters at the corresponding time under the same operating conditions, resulting in a first load prediction curve. The historical load curve under the same operating conditions is input into a second prediction model to obtain a second load prediction curve, representing the continuous production characteristics of the pumping unit well. Based on the first and second load prediction curves, a target prediction curve for the pumping unit well is obtained. The first load prediction curve reflects the changes in the production state of the pumping unit well at a specific time, while the second load prediction curve, based on the historical load curve, derives the long-term trend using continuous production characteristics. The combination of the two prediction curves effectively reduces the error caused by data limitations in a single prediction model, improving the accuracy and reliability of the prediction.

[0057] Based on this, embodiments of this application provide a method for predicting the load of a pumping unit well, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the load prediction method for pumping wells in this application.

[0058] In this embodiment, the load prediction method for the pumping well includes steps S10 to S40:

[0059] Step S10: Obtain the static parameters and load parameters of the pumping well at different times under the same operating conditions, and obtain the mapping relationship between the static parameters and load parameters through the first prediction model.

[0060] It should be noted that the static parameters of a pumping unit well under the same operating conditions at different times may include the pumping unit type, well dynamometer card, well operating conditions, well stroke, number of strokes, and well production fluid volume, etc. The load parameters can be the load parameters corresponding to the static parameters. It can be understood that the mapping relationship is used to represent the discrete characteristics of pumping unit well production, and the first prediction model can be a model used to establish the mapping relationship between static parameters and load parameters. For example, the first prediction model can be a fully connected neural network.

[0061] Step S20: Based on the mapping relationship and the static parameters of the pumping unit well at different times under the same working condition, the first prediction model is used to predict the pumping unit load parameters at the corresponding time under the same working condition, and the first load prediction curve is obtained.

[0062] It should be noted that the first load prediction curve can reflect the discrete characteristics of pumping well production. Under the same operating conditions, the load parameters at each moment are calculated using static parameters at different times, and the first load prediction curve is plotted based on the predicted load parameters at each moment.

[0063] Step S30: Input the historical load curve under the same operating condition into the second prediction model to obtain the second load prediction curve.

[0064] It should be noted that the historical load curve can describe the time series of load parameters recorded by the pumping unit well during its historical operation, and can represent the continuous production characteristics of the pumping unit well. The second prediction model can be a model that predicts future load changes based on the historical load curve. For example, the second prediction model can adopt a Long Short-Term Memory (LSTM) network or other time series models, and the second load prediction curve can be a time series of load parameters predicted based on the historical load curve and the second prediction model.

[0065] Step S40: Based on the first load prediction curve and the second load prediction curve, obtain the target prediction curve for the pumping well.

[0066] It should be noted that the target forecast curve can be understood as the final load forecast curve that combines the first load forecast curve (discrete characteristics) and the second load forecast curve (continuous characteristics).

[0067] In this embodiment, static parameters and load parameters of the pumping unit well at different times under the same operating condition are obtained. A first prediction model is used to obtain the mapping relationship between the static parameters and the load parameters, representing the discrete production characteristics of the pumping unit well. Based on the mapping relationship and the static parameters of the pumping unit well at different times under the same operating condition, the first prediction model predicts the pumping unit load parameters at the corresponding time under the same operating condition, obtaining a first load prediction curve. The historical load curve under the same operating condition is input into a second prediction model to obtain a second load prediction curve, representing the continuous production characteristics of the pumping unit well. Based on the first and second load prediction curves, the target prediction curve for the pumping unit well is obtained. The first load prediction curve reflects the changes in the production state of the pumping unit well at a specific time, while the second load prediction curve, based on the historical load curve, derives the long-term trend using continuous production characteristics. The combination of the two prediction curves effectively reduces the error caused by data limitations in a single prediction model, improving the accuracy and reliability of the prediction.

[0068] Reference Figure 2 , Figure 2 This is a flowchart illustrating the second embodiment of the load prediction method for pumping wells in this application, based on the above. Figure 1 The first embodiment shown presents a second embodiment of the load prediction method for pumping wells according to this application.

[0069] In the second embodiment, step S10 includes:

[0070] Step S101: Preprocess the static parameters and load parameters of the pumping well at different times under the same working conditions, and input the preprocessed data into the input layer of the first prediction model to obtain the first dataset.

[0071] It should be noted that preprocessing can include cleaning, normalization, and format conversion. The first dataset can be preprocessed static parameter and load parameter data, used to train and test the first prediction model.

[0072] Step S102: Input the first dataset into the hidden layer and extract the mapping relationship between static parameters and load parameters through a multi-layer neural network structure.

[0073] It should be noted that the hidden layer can be an intermediate layer of the first prediction model, consisting of multiple neurons. In the hidden layer, the relationship between the static parameters and the load parameters can be extracted into high-dimensional features through nonlinear transformation.

[0074] Step S103: Output the mapping relationship between static parameters and load parameters through the output layer.

[0075] It should be noted that the output layer is the final layer of the first prediction model, used to generate the model's output results. It can be understood that when the first dataset is input into the first prediction model, the corresponding output results are the mapping relationship between static parameters and load parameters.

[0076] In this embodiment, by inputting preprocessed data into the first prediction model, data quality and model accuracy can be improved. Through a multi-layered neural network with hidden layers, deep-level features of the data can be captured and transformed into features that the model can understand and learn. By outputting the mapping relationship between static parameters and load parameters in a clear numerical form through the output layer, it is possible to clearly reflect how static parameters affect the load at each moment, providing strong support for load prediction in discrete states.

[0077] Reference Figure 3 , Figure 3 This is a flowchart illustrating the third embodiment of the load prediction method for pumping wells in this application, based on the above. Figure 2 The second embodiment shown presents a third embodiment of the load prediction method for pumping wells according to this application.

[0078] In the third embodiment, step S20 includes:

[0079] Step S201: Based on the mapping relationship and the static parameters of the pumping unit well at different times under the same operating conditions, the feature information of the static parameters of the pumping unit well at different times under the same operating conditions is extracted through the hidden layer.

[0080] It should be noted that feature information can be understood as the more important information in the input data (such as static parameters) extracted through the hidden layers of a neural network that can effectively represent the characteristics of the data.

[0081] Step S202: The feature information is nonlinearly mapped using an activation function to generate predicted load parameters corresponding to the feature information.

[0082] It should be noted that activation functions are used to perform nonlinear transformations on the feature information extracted from the hidden layer. For example, activation functions may include ReLU (Rectified Linear Unit), Sigmoid, Tanh, etc. The predicted load parameters can be understood as the load values ​​of the pumping unit well under specific operating conditions predicted by a neural network. Through nonlinear mapping, the neural network transforms static parameters and mapping relationships into corresponding predicted load values, which reflect the operating status of the pumping unit under different conditions.

[0083] Step S203: Based on the predicted load parameters, obtain the first load prediction curve under the same operating condition.

[0084] It should be noted that the load forecast value calculated by the neural network is the result of the model output under a given input (such as static parameters). The first load forecast curve is a curve that shows the predicted load parameter value as time or other variables change under a specific operating condition.

[0085] In this embodiment, feature information of the static parameters of the pumping unit well at different times under the same operating conditions is extracted through a hidden layer based on static parameters. Furthermore, a deep relationship between the static parameters and the load is learned through nonlinear mapping, effectively capturing the impact of operating condition changes on the load and thus improving the accuracy of load prediction. Using a multi-layer neural network structure and activation functions to perform nonlinear mapping on the feature information enhances the model's performance under complex operating conditions, enabling it to better adapt to the load prediction needs under different production conditions and improving the predictive versatility and scalability.

[0086] Reference Figure 4 , Figure 4 This is a flowchart illustrating the fourth embodiment of the load prediction method for pumping wells in this application, based on the above. Figure 3 The third embodiment shown presents a fourth embodiment of the load prediction method for pumping wells according to this application.

[0087] In the fourth embodiment, step S30 includes:

[0088] Step S301: Input the historical load curve under the same working condition into the input layer of the second prediction model to obtain the second dataset.

[0089] It should be noted that historical load curves refer to time-series data of pumping wells under the same operating conditions over a past period, typically recorded by monitoring equipment. Time-series data reflects the load variation trend of pumping wells during continuous production. The second dataset can be preprocessed load variation data. For example, load variation data for 5 minutes of continuous production from different types of wells under different operating conditions can be selected. After preprocessing (0-1 normalization), this data is input into the input layer of the second prediction model to obtain 5-minute time-series data of continuous production from the pumping wells. The second prediction model can be a model for processing time-series data, such as a Long Short-Term Memory (LSTM) network or a Recurrent Neural Network (RNN), which can be used to capture the time dependence in historical load curves and predict future load changes.

[0090] It should be noted that, because load curves often exhibit a bimodal characteristic (e.g., Figure 5 As shown in the diagram, it cannot be simply assumed that the end of a downward trend is the dead point of the downstroke, or the end of an upward trend is the dead point of the upstroke. Therefore, it is necessary to combine the overall and local changes of the curve to make a judgment. The Long Short-Term Memory Recurrent Neural Network (LSTM) can summarize the long-term and short-term trends of load changes. Therefore, the LSTM network can be selected as the main network of the second prediction model.

[0091] Step S302: Input the second dataset into the hidden layer of the second prediction model to generate load prediction values ​​corresponding to the historical time series.

[0092] It should be noted that the hidden layer can be understood as the core part of the second prediction model, used to extract key features from the time series data. For example, the second dataset is input into the hidden layer of the second prediction model. According to the time step processing mechanism, the data of each time step is associated with the features of the previous time step, thereby capturing the short-term changes (local characteristics) and long-term dependencies (global characteristics) of the time series. The feature information output by the hidden layer is passed to the prediction module (which can be a fully connected layer or a linear regression layer) to generate the load prediction value for each time step.

[0093] Step S303: Reconstruct the load forecast values ​​of multiple time steps to obtain the second load forecast curve.

[0094] For example, these predicted values ​​can be combined in chronological order to form a complete predicted load curve. By determining the sequence order of time steps, it is ensured that the predicted values ​​correspond one-to-one with the actual time points, and the curve is smoothed to eliminate short-term noise. For example, the historical load curves before and after the smoothing operation are divided into... Figure 6 and Figure 7 As shown, it is understandable that the historical load curve is as follows: Figure 8As shown, the blue curve represents the historical load curve, and the orange curve represents the second load forecast curve. It can be understood that the period between the two red dots is one cycle of the historical load curve, and the blue dot can be understood as the starting point for calculating the cycle of the second load forecast curve.

[0095] In this embodiment, the periodic load characteristics of the pumping well can be accurately captured by the time series characteristics and curve reconstruction of the second prediction model. The reconstructed curve can show the dynamic change process of the load, which makes it easier to understand the operating status of the pumping well.

[0096] In one implementation, based on the fourth embodiment described above, step S302 includes: inputting the second dataset into a long short-term memory unit, and extracting historical load information through a memory gate, input gate, and forget gate mechanism; inputting the historical load information into the hidden layer of the second prediction model, and extracting time-dependent features through multiple long short-term memory units, the time-dependent features capturing the continuous production characteristics of pumping wells; and inputting the time-dependent features into a fully connected layer to generate load prediction values ​​corresponding to the historical time series.

[0097] It should be noted that the forget gate can decide which past time step information to retain or discard, the input gate can update the memory cell state using the data of the current time step, and the memory gate can retain long-term dependency features throughout the entire time series.

[0098] In this embodiment, by inputting the second dataset into the Long Short-Term Memory (LSTM) unit to extract historical load information, the overall trend and local changes of the curve can be comprehensively considered, ensuring the accuracy of the historical load information. By capturing time-dependent features, the model can provide high-precision prediction results, providing a reliable basis for production optimization and equipment scheduling. It is not only suitable for load prediction under a single operating condition, but can also be flexibly extended to the prediction of complex load curves under various production conditions.

[0099] In one embodiment, based on the above embodiments and implementation methods, step S40 includes: averaging and superimposing the predicted values ​​at corresponding time points of the first load prediction curve and the second load prediction curve to obtain the target prediction value of the pumping well; and obtaining the target prediction curve of the pumping well based on the target prediction value of the pumping well.

[0100] For example, by averaging and superimposing the predicted values ​​at corresponding time points of the first load prediction curve and the second load prediction curve, the target predicted value of the pumping well can be expressed as formula (1):

[0101]

[0102] In formula (1), P(t) represents the target predicted value of the pumping well, P1(t) represents the predicted value at the corresponding time point of the first load prediction curve, and P2(t) represents the predicted value at the corresponding time point of the second load prediction curve.

[0103] It should be noted that, in addition to averaging the predicted values ​​at corresponding time points of the first and second load prediction curves, the target predicted value can also be calculated by weighted superposition of the predicted values ​​at corresponding time points of the first and second load prediction curves. For example, by weighted superposition of the predicted values ​​at corresponding time points of the first and second load prediction curves, the target predicted value of the pumping unit well can be expressed as formula (2):

[0104] P(t)=ω1P1(t)+ω2P2(t) (2)

[0105] In formula (2), P(t) represents the target predicted value of the pumping well, ω1 represents the weight of the first load prediction curve, P1(t) represents the predicted value of the corresponding time point of the first load prediction curve, ω2 represents the weight of the first load prediction curve, and P2(t) represents the predicted value of the corresponding time point of the second load prediction curve.

[0106] In this embodiment, by fusing discrete characteristics (first load prediction curve) and continuous characteristics (second load prediction curve), the target prediction curve can more comprehensively reflect the actual production characteristics of the pumping well. The average superposition method can integrate the information of the first and second load prediction curves, ensuring that the target prediction curve has a certain degree of robustness.

[0107] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the load prediction method for pumping wells in this application. Any simple modifications based on this technical concept are within the protection scope of this application.

[0108] This application also provides a load prediction device for pumping wells; please refer to [reference needed]. Figure 9 The load prediction device for the pumping well includes:

[0109] The mapping relationship acquisition module 10 is used to acquire the static parameters and load parameters of the pumping well at different times under the same working condition, and obtain the mapping relationship between the static parameters and load parameters through the first prediction model. The mapping relationship is used to represent the discrete production characteristics of the pumping well.

[0110] The first prediction curve module 20 is used to predict the pumping unit load parameters at the corresponding time under the same working condition based on the mapping relationship and the static parameters of the pumping unit well at different times under the same working condition, and obtain the first load prediction curve.

[0111] The second prediction curve module 30 is used to input the historical load curve under the same working condition into the second prediction model to obtain the second load prediction curve. The historical load curve is used to represent the continuous production characteristics of the pumping well.

[0112] The target prediction curve module 40 is used to obtain the target prediction curve of the pumping well based on the first load prediction curve and the second load prediction curve.

[0113] The load prediction device for pumping unit wells provided in this application, employing the load prediction method for pumping unit wells in the above embodiments, can solve the technical problem of low prediction accuracy of pumping unit load. Compared with the prior art, the beneficial effects of the load prediction device for pumping unit wells provided in this application are the same as the beneficial effects of the load prediction method for pumping unit wells provided in the above embodiments, and other technical features in the load prediction device for pumping unit wells are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0114] This application provides a load prediction device for a pumping well. The load prediction device for a pumping well includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the load prediction method for the pumping well in the above embodiment 1.

[0115] The following is for reference. Figure 10 This document illustrates a structural schematic diagram of a load prediction device suitable for implementing the embodiments of this application for oil pumping wells. The load prediction device for oil pumping wells in the embodiments of this application may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital radio receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), vehicle terminals (e.g., vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 10 The load prediction device for the pumping well shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.

[0116] like Figure 10As shown, the load prediction device for a pumping unit well may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM) 1004. The RAM 1004 also stores various programs and data required for the operation of the pumping unit well load prediction device. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the load prediction equipment of the pumping well to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 10 Load prediction equipment for pumping wells with various systems is shown; however, it should be understood that implementation or possession of all the systems shown is not required. More or fewer systems may be implemented alternatively.

[0117] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0118] The load prediction device for pumping unit wells provided in this application, employing the load prediction method for pumping unit wells in the above embodiments, can solve the technical problem of low prediction accuracy of pumping unit load. Compared with the prior art, the beneficial effects of the load prediction device for pumping unit wells provided in this application are the same as the beneficial effects of the load prediction method for pumping unit wells provided in the above embodiments, and other technical features in the load prediction device for pumping unit wells are the same as those disclosed in the method of the previous embodiment, and will not be repeated here.

[0119] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0120] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0121] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the load prediction method for pumping wells in the above embodiments.

[0122] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0123] The aforementioned computer-readable storage medium may be included in the load prediction device of the pumping well; or it may exist independently and not be installed in the load prediction device of the pumping well.

[0124] The aforementioned computer-readable storage medium carries one or more programs. When these programs are executed by the load prediction device of the pumping unit well, the load prediction device of the pumping unit well performs the following actions: acquires the static parameters and load parameters of the pumping unit well at different times under the same operating condition; obtains a mapping relationship between the static parameters and the load parameters through a first prediction model, wherein the mapping relationship is used to represent the discrete production characteristics of the pumping unit well; based on the mapping relationship and the static parameters of the pumping unit well at different times under the same operating condition, predicts the pumping unit load parameters at the corresponding time under the same operating condition through the first prediction model, thereby obtaining a first load prediction curve; inputs the historical load curve under the same operating condition into a second prediction model to obtain a second load prediction curve, wherein the historical load curve is used to represent the continuous production characteristics of the pumping unit well; and obtains the target prediction curve of the pumping unit well based on the first load prediction curve and the second load prediction curve.

[0125] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a Local Area Network (LAN) or a Wide Area Network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0126] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0127] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0128] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described load prediction method for pumping wells, thereby solving the technical problem of low prediction accuracy for pumping unit loads. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the load prediction method for pumping wells provided in the above embodiments, and will not be repeated here.

[0129] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described method for predicting the load of a pumping well.

[0130] The computer program product provided in this application can solve the technical problem of low prediction accuracy of pumping unit load. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the pumping unit well load prediction method provided in the above embodiments, and will not be repeated here.

[0131] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A method for predicting the load of a pumping well, characterized in that, The method includes: The static parameters and load parameters of the pumping unit well at different times under the same operating conditions are obtained. The mapping relationship between the static parameters and the load parameters is obtained through the first prediction model. The mapping relationship is used to represent the discrete production characteristics of the pumping unit well. Based on the mapping relationship and the static parameters of the pumping unit well at different times under the same working condition, the pumping unit load parameters at the corresponding time under the same working condition are predicted by the first prediction model to obtain the first load prediction curve. The historical load curve under the same operating condition is input into the second prediction model to obtain the second load prediction curve, which is used to represent the continuous production characteristics of the pumping well. Based on the first load prediction curve and the second load prediction curve, the target prediction curve for the pumping well is obtained.

2. The method as described in claim 1, characterized in that, The step of obtaining the static parameters and load parameters of the pumping unit well at different times under the same operating conditions, and obtaining the mapping relationship between the static parameters and the load parameters, includes: The static parameters and load parameters of the pumping well under the same working condition at different times are preprocessed, and the preprocessed data is input into the input layer of the first prediction model to obtain the first dataset; The first dataset is input into the hidden layer, and the mapping relationship between static parameters and load parameters is extracted through a multi-layer neural network structure; The mapping relationship between the static parameters and the load parameters is output through the output layer.

3. The method as described in claim 2, characterized in that, The step of predicting the pumping unit load parameters at corresponding times under the same operating condition using a first prediction model based on the mapping relationship and the static parameters of the pumping unit well at different times under the same operating condition, and obtaining the first load prediction curve, includes: Based on the mapping relationship and the static parameters of the pumping unit well at different times under the same working condition, feature information corresponding to the static parameters of the pumping unit well at different times under the same working condition is extracted through the hidden layer. The feature information is nonlinearly mapped by an activation function to generate predicted load parameters corresponding to the feature information. Based on the predicted load parameters, the first load prediction curve under the same operating condition is obtained.

4. The method as described in claim 1, characterized in that, The step of inputting the historical load curve under the same operating condition into the second prediction model to obtain the second load prediction curve includes: The historical load curves under the same operating conditions are input into the input layer of the second prediction model to obtain the second dataset; The second dataset is input into the hidden layer of the second prediction model to generate load forecast values ​​corresponding to the historical time series; The load forecast values ​​at multiple time steps are reconstructed to obtain a second load forecast curve.

5. The method as described in claim 4, characterized in that, The step of inputting the second dataset into the hidden layer of the second prediction model to generate load forecast values ​​corresponding to historical time series includes: The second dataset is input into the long short-term memory unit, and historical load information is extracted through the memory gate, input gate and forget gate mechanisms; The historical load information is input into the hidden layer of the second prediction model, and time-dependent features are extracted through multi-layer long short-term memory units. The time-dependent features capture the continuous production characteristics of the pumping well. The time-dependent features are input into a fully connected layer to generate load forecast values ​​corresponding to historical time series.

6. The method according to any one of claims 1 to 5, characterized in that, The step of obtaining the target prediction curve of the pumping unit well based on the first load prediction curve and the second load prediction curve includes: The target predicted value of the pumping well is obtained by averaging and superimposing the predicted values ​​at corresponding time points of the first load prediction curve and the second load prediction curve. Based on the target predicted value of the pumping well, the target prediction curve of the pumping well is obtained.

7. A load prediction device for oil pumping wells, characterized in that, The device includes: The mapping relationship acquisition module is used to acquire the static parameters and load parameters of the pumping well at different times under the same working condition, and obtain the mapping relationship between the static parameters and load parameters through the first prediction model. The mapping relationship is used to represent the discrete production characteristics of the pumping well. The first prediction curve module is used to predict the pumping unit load parameters at the corresponding time under the same working condition based on the mapping relationship and the static parameters of the pumping unit well at different times under the same working condition, and obtain the first load prediction curve. The second prediction curve module is used to input the historical load curve under the same operating condition into the second prediction model to obtain the second load prediction curve. The historical load curve is used to represent the continuous production characteristics of the pumping well. The target prediction curve module is used to obtain the target prediction curve of the pumping well based on the first load prediction curve and the second load prediction curve.

8. A load prediction device for oil pumping wells, characterized in that, The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the load prediction method for pumping wells as described in any one of claims 1 to 6.

9. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the load prediction method for pumping wells as described in any one of claims 1 to 6.

10. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the steps of the load prediction method for pumping wells as described in any one of claims 1 to 6.

Images