An intelligent analysis method and device for pumping wells

Abstract

This article relates to the technical field of intelligent production of oil wells, and provides an intelligent analysis method and device for a pumping well, wherein the method includes: collecting electrical parameter curves of the pumping well; performing frequency domain transformation on the collected electrical parameter curves to obtain the frequency spectrum characteristics of the electrical parameter curves; and calculating the working condition information of the pumping well using a working condition diagnosis model according to the frequency spectrum characteristics of the electrical parameter curves, wherein the working condition diagnosis model is trained by the frequency spectrum characteristics of the historical electrical parameter curves under different working conditions of the pumping well. This article uses the frequency spectrum characteristics of the electrical parameter curves to identify the working conditions, thereby eliminating the process of identifying the upper and lower dead points of the electrical parameter curves, eliminating the interference of the balance of the pumping well on the characteristics of the electrical parameter curves, and improving the accuracy of working condition diagnosis.

Description

Technical Field

[0001] This document relates to the technical field of intelligent oil well production, and particularly to an intelligent analysis method and device for pumping wells. Background Art

[0002] When building the Internet of Things for pumping wells, the conventional configuration includes an electrical parameter tester and a dynamometer card tester. The electrical parameters collected by the electrical parameter tester are mainly used to reflect the working conditions of surface equipment, and the dynamometer card generated by the dynamometer card tester is used to reflect the working conditions of downhole equipment. Since the current Internet of Things for oil wells includes both the relatively high-cost dynamometer card mode and the relatively low-cost electrical parameter mode, the intelligent method for analyzing the production of pumping wells based on the Internet of Things needs to consider this situation, requiring the ability to diagnose and optimize pumping wells based on both the dynamometer card and electrical parameters.

[0003] In the prior art, the dynamometer card can directly perform working condition diagnosis. Therefore, the diagnosis and optimization methods based on the dynamometer card are relatively mature. However, the dynamometer card tester has some drawbacks, such as: high investment cost, easy drift and distortion caused by the mechanical alternating load endured by the dynamometer card tester when installed at the polished rod for a long time, and high calibration and maintenance costs. These drawbacks will affect the large-scale application of the Internet of Things for oil wells.

[0004] In the prior art, when using the electrical parameter curve to perform working condition diagnosis, it is necessary to first identify the upper and lower dead points in the electrical parameter curve, and then directly perform working condition diagnosis based on the electrical parameter curve between the upper and lower dead points. The existing methods for performing working condition diagnosis using the electrical parameter curve have the following problems: 1) The identification of the upper and lower dead points in the electrical parameter curve is usually achieved using an angular displacement sensor, which will increase the cost; 2) Since there are many energy transmission links from the ground to the downhole for electrical parameters and there are many influencing factors for electrical parameters, compared with the dynamometer card, the characteristics of the electrical parameter curve are more complex, and it is difficult to identify the upper and lower dead points of the electrical parameter curve. This makes the existing method for directly performing working condition diagnosis using the electrical parameter curve have problems of low accuracy and difficulty in implementation. Summary of the Invention

[0005] This document is used to solve the problems of low accuracy and difficulty in implementation existing in the prior art for the diagnosis method based on electrical parameters.

[0006] To solve the above technical problems, the first aspect of this document provides an intelligent analysis method for pumping wells, including:

[0007] Collect the electrical parameter curve of the pumping well;

[0008] Perform frequency domain transformation on the collected electrical parameter curve to obtain the spectral characteristics of the electrical parameter curve;

[0009] According to the spectral characteristics of the electric parameter curve, the working condition information of the pumping unit well is calculated by using a working condition diagnosis model, wherein the working condition diagnosis model is trained by the spectral characteristics of the historical electric parameter curves under different working conditions of the pumping unit well.

[0010] In a further embodiment of the present invention, the intelligent analysis method for a pumping unit well further includes:

[0011] Calculate an electric parameter difference curve according to the collected electric parameter curve and the initial electric parameter curve of the pumping unit well;

[0012] Calculate a polished rod load difference curve according to the electric parameter difference curve;

[0013] Superimpose the polished rod load difference curve onto the initial indicator diagram to determine the polished rod displacement value;

[0014] Draw an indicator diagram curve according to the polished rod displacement value and the polished rod load value obtained after superposition;

[0015] Analyze the working condition information of the pumping unit well by using the indicator diagram curve;

[0016] Judge whether the working condition information of the pumping unit well calculated by using the working condition diagnosis model is consistent with the working condition information of the pumping unit well analyzed by using the indicator diagram curve. If they are consistent, output the working condition information; if they are not consistent, send an alarm message.

[0017] In a further embodiment of the present invention, the intelligent analysis method for a pumping unit well further includes:

[0018] Query a first optimization database according to the working condition of the pumping unit well to obtain an optimization method, wherein the corresponding relationship between the working condition and the optimization method is stored in the first optimization database;

[0019] Control the pumping unit well according to the queried optimization method.

[0020] In a further embodiment of the present invention, the intelligent analysis method for a pumping unit well further includes:

[0021] Collect the production data of the pumping unit well;

[0022] Calculate the production of the pumping unit well by using a production calculation model according to the collected production data and the spectral characteristics of the electric parameter curve;

[0023] Wherein, the production calculation model is trained by the historical production data corresponding to each production of the pumping unit well and the spectral characteristics of the historical electric parameter curves.

[0024] In a further embodiment of the present invention, the intelligent analysis method for a pumping unit well further includes:

[0025] Query the second optimization database according to the working conditions of the pumping unit well and the production of the pumping unit well to obtain an optimization method, where the second optimization database stores the corresponding relationship between the working conditions, production, and optimization methods;

[0026] Control the pumping unit well according to the queried optimization method.

[0027] In a further embodiment of this article, the process of establishing the working condition diagnosis model includes:

[0028] Collect the electrical parameter curves of the pumping unit well under different working conditions;

[0029] Perform frequency domain transformation on the collected electrical parameter curves to obtain the spectral characteristics of each electrical parameter curve;

[0030] Take the working condition information of the pumping unit well as the output, and the spectral characteristics of the electrical parameter curve under the corresponding working condition as the input, and train to obtain a working condition diagnosis model.

[0031] In a further embodiment of this article, taking the working condition information of the pumping unit well as the output, and the spectral characteristics of the electrical parameter curve under the corresponding working condition as the input, and training to obtain a working condition diagnosis model, includes:

[0032] Take the working condition or the working condition probability value of the pumping unit well as the output, and the spectral characteristics of the electrical parameter curve under the corresponding working condition as the input, and train using a machine learning model and a deep learning model respectively to obtain a first diagnosis model and a second diagnosis model;

[0033] The working condition diagnosis model is composed of the first diagnosis model and the second diagnosis model.

[0034] In a further embodiment of this article, the working condition diagnosis model is composed of the first diagnosis model and the second diagnosis model, including:

[0035] Perform weighted processing on the first diagnosis model and the second diagnosis model to obtain the working condition diagnosis model.

[0036] In a further embodiment of this article, the machine learning model includes: one of support vector machine and K-nearest neighbor method; the deep learning model includes: one of fully connected network model and recurrent neural network model.

[0037] In a further embodiment of this article, after collecting the electrical parameter curves of the pumping unit well under different working conditions, it further includes:

[0038] Perform preprocessing on the electrical parameter curves of the pumping unit well under different working conditions.

[0039] In a further embodiment of this article, performing preprocessing on the electrical parameter curves of the pumping unit well under different working conditions includes:

[0040] Smooth, denoise, and standardize the electrical parameter curves of the pumping unit well under different working conditions respectively.

[0041] In a further embodiment of this article, according to the electrical parameter difference curve, calculate the polished rod load difference curve, including calculating the polished rod load difference curve using the following formula:

[0042]

[0043] Where, ΔF(i) is the polished rod load difference of the pumping unit well at the i-th moment within a single pumping unit stroke, with the unit of kN; K is the motor efficiency, dimensionless; ΔP(i) is the difference in the input active power of the pumping unit at the i-th moment within a single pumping unit stroke; η1 is the belt drive efficiency; η2 is the reducer drive efficiency; i MB is the transmission ratio from the output shaft of the pumping unit to the output shaft of the reducer; n m is the motor speed, with the unit of rpm; TF is the torque factor, dimensionless.

[0044] The second aspect of this article provides an intelligent analysis device for a pumping unit well, including:

[0045] A first acquisition module for acquiring the electrical parameter curve of the pumping unit well;

[0046] A frequency spectrum transformation module for performing frequency domain transformation on the acquired electrical parameter curve to obtain the frequency spectrum characteristics of the electrical parameter curve;

[0047] A first analysis module for calculating the working condition information of the pumping unit well according to the frequency spectrum characteristics of the electrical parameter curve by using a working condition diagnosis model, where the working condition diagnosis model is trained by the frequency spectrum characteristics of the historical electrical parameter curves of the pumping unit well under different working conditions.

[0048] In a further embodiment of this article, the intelligent analysis device for a pumping unit well further includes:

[0049] A calculation module for calculating the electrical parameter difference curve according to the acquired electrical parameter curve and the initial electrical parameter curve of the pumping unit well; calculating the polished rod load difference curve according to the electrical parameter difference curve; superimposing the polished rod load difference curve onto the initial indicator diagram to determine the polished rod displacement; and drawing an indicator diagram curve according to the polished rod displacement value and the polished rod load value obtained after superposition;

[0050] A second analysis module for analyzing the working condition information of the pumping unit well by using the indicator diagram curve;

[0051] A judgment module is used to determine whether the working condition information of the pumping well calculated by using the working condition diagnosis model is consistent with the working condition information of the pumping well analyzed from the dynamometer card curve. If they are consistent, the working condition information is output; if not, an alarm message is sent.

[0052] In the third aspect of this article, a computer device is provided, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the intelligent analysis method for pumping wells described in the foregoing embodiments is implemented.

[0053] In the fourth aspect of this article, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the intelligent analysis method for pumping wells described in the foregoing embodiments is implemented.

[0054] The intelligent analysis method and device for pumping wells provided in this article can save the process of identifying the top and bottom dead points of the electrical parameter curve, eliminate the interference of the balance degree of the pumping well on the characteristics of the electrical parameter curve, and improve the accuracy of working condition diagnosis by using the spectral characteristics of the electrical parameter curve for working condition identification.

[0055] To make the above and other purposes, features, and advantages of this article more obvious and understandable, the following specific preferred embodiments are given in conjunction with the accompanying drawings and described in detail as follows. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] To more clearly illustrate the technical solutions in the embodiments of this article or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of this article. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0057] Figure 1A Shows the first flowchart of the intelligent analysis method for pumping wells in the embodiments of this article;

[0058] Figure 1B Shows the second flowchart of the intelligent analysis method for pumping wells in the embodiments of this article;

[0059] Figure 2 Shows the third flowchart of the intelligent analysis method for pumping wells in the embodiments of this article;

[0060] Figure 3 Shows the fourth flowchart of the intelligent analysis method for pumping wells in the embodiments of this article;

[0061] Figure 4 Shows the fifth flowchart of the intelligent analysis method for pumping wells in the embodiments of this article;

[0062] Figure 5 Shows the first flowchart of the working condition diagnosis model establishment process of the embodiments herein;

[0063] Figure 6 Shows the second flowchart of the working condition diagnosis model establishment process of the embodiments herein;

[0064] Figure 7 Shows the second flowchart of the working condition diagnosis model establishment process of the embodiments herein;

[0065] Figure 8A Shows the first structure diagram of the intelligent analysis device for pumping unit wells of the embodiments herein;

[0066] Figure 8B Shows the second structure diagram of the intelligent analysis device for pumping unit wells of the embodiments herein;

[0067] Figure 9 Shows the third structure diagram of the intelligent analysis device for pumping unit wells of the embodiments herein;

[0068] Figure 10 Shows the fourth structure diagram of the intelligent analysis device for pumping unit wells of the embodiments herein;

[0069] Figure 11 Shows the fifth structure diagram of the intelligent analysis device for pumping unit wells of the embodiments herein;

[0070] Figure 12 Shows the flowchart of the intelligent analysis method for pumping unit wells of the specific embodiments herein;

[0071] Figure 13 Shows the structure diagram of the computer device of the embodiments herein.

[0072] Explanation of the reference numerals in the drawings:

[0073] 810, First acquisition module;

[0074] 820, Spectrum transformation module;

[0075] 830, First analysis module;

[0076] 840, Calculation module;

[0077] 850, Second analysis module;

[0078] 860, Judgment module;

[0079] 870, First optimization module;

[0080] 880, Second acquisition module;

[0081] 890. Third analysis module;

[0082] 891. Second optimization module;

[0083] 1302. Computer device;

[0084] 1304. Processor;

[0085] 1306. Memory;

[0086] 1308. Driving mechanism;

[0087] 1310. Input / output module;

[0088] 1312. Input device;

[0089] 1314. Output device;

[0090] 1316. Presentation device;

[0091] 1318. Graphical user interface;

[0092] 1320. Network interface;

[0093] 1322. Communication link;

[0094] 1324. Communication bus. Detailed implementation manners

[0095] Next, the technical solutions in the embodiments of this article will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of this article. Obviously, the described embodiments are only a part of the embodiments of this article, rather than all the embodiments. Based on the embodiments in this article, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the scope of protection of this article.

[0096] As Figure 1A shown, Figure 1A a flowchart of the intelligent analysis method for pumping wells in the embodiments of this article is shown. In this embodiment, by using the spectral characteristics of the electrical parameter curve for working condition identification, the process of identifying the top and bottom dead points of the electrical parameter curve can be omitted, the interference of the balance degree of the pumping well on the characteristics of the electrical parameter curve can be eliminated, and the accuracy of working condition diagnosis can be improved. Specifically, the intelligent analysis method for pumping wells includes:

[0097] Step 110: Collect the electrical parameter curve of the pumping well.

[0098] In specific implementation, by connecting to the oilfield Internet of Things of the pumping well, the electrical parameter curve of the pumping well is collected from the electrical parameter tester. The electrical parameter curve includes three-phase current, voltage, and active power. To ensure the integrity of information, the collected electrical parameter curve of the pumping well includes at least the information of a complete stroke cycle.

[0099] During implementation, other production and test historical data such as the stroke, pumping frequency, pump depth, pump diameter, liquid level, and liquid production of the pumping unit well can also be collected and used as samples for subsequent big data deep learning.

[0100] Step 120: Perform a frequency-domain transformation on the collected electrical parameter curve to obtain the spectral characteristics of the electrical parameter curve.

[0101] During specific implementation, a discrete Fourier transform can be used to perform a frequency-domain transformation on the collected electrical parameter curve, or a wavelet transform method can also be used. This article does not specifically limit the frequency-domain transformation method. The spectral characteristics of the electrical parameter curve are a characteristic sequence composed of the spectral energy amplitudes in different frequency domains. The following takes the discrete Fourier transform as an example to illustrate the transformation process. The discrete Fourier transform is used to obtain the frequency-domain characteristics of the electrical parameters. The discrete Fourier transform of a finite-length sequence is:

[0102]

[0103] Then the real part is:

[0104]

[0105] The imaginary part is:

[0106]

[0107] Taking the positive square root value of the sum of the squares of the real part and the imaginary part is the spectral characteristic of the k-th frequency domain part.

[0108] Using the spectral characteristics to diagnose and analyze the electrical parameters of the pumping unit well has the following advantages:

[0109] 1) The spectral characteristics of the electrical parameter curve are independent of the starting point of the electrical parameter curve, so there is no need to use an angular displacement sensor to identify the top and bottom dead points in the electrical parameter curve during the construction of the Internet of Things for pumping unit wells.

[0110] 2) When the balance degree of the pumping unit well changes, the spectral characteristics of the electrical parameter curve will change significantly (the fundamental frequency energy amplitude in the spectral characteristics changes, and other frequency domains do not change). The spectral characteristics of the electrical parameter curve can eliminate the interference of the balance degree of the pumping unit well on the characteristics of the electrical parameter curve.

[0111] 3) Under different severity conditions of different working conditions of the pumping unit well, the variation laws and degrees of the spectral characteristics of the electrical parameter curve are also different. It can be determined that the spectral characteristics of the electrical parameter curve have good discrimination for the working conditions of the pumping unit well. For example, under the condition of rod breakage, the energy amplitude of the fundamental frequency of the electrical parameter curve will increase significantly, and the energy amplitudes of other frequencies will suddenly decrease. Moreover, the closer the rod breakage position is to the wellhead, the more significant this change trend is, which can be used as an important feature of the rod breakage condition. Another example is that under the condition of wax deposition, the spectral characteristics mainly used to distinguish the working conditions are: first, the DC component and the amplitude in the double-frequency domain increase with the increase of the wax deposition degree; second, the high-frequency signal completely disappears. Another example is that under the condition of leakage of the traveling valve, the spectral characteristics mainly used to distinguish the working conditions are: the energy amplitudes of the fundamental frequency and the triple frequency increase with the increase of the leakage degree, and the energy amplitudes in the frequency domain of the double frequency and above the quadruple frequency decrease with the increase of the leakage degree.

[0112] By converting the electrical parameter curve into the spectral characteristics of the electrical parameter curve in step 120 and using the spectral characteristics for the working condition diagnosis of the pumping unit well, the accuracy of the working condition diagnosis can be improved, and the steps of identifying the top and bottom dead points of the electrical parameter curve can be omitted.

[0113] Step 130, according to the spectral characteristics of the electrical parameter curve obtained in step 120, calculate the working condition information of the pumping unit well by using the working condition diagnosis model.

[0114] Among them, the working condition diagnosis model is trained by the spectral characteristics of the electrical parameter curves under different working conditions of the pumping unit well collected historically. The determination process of the working condition diagnosis model will be described in detail in the subsequent embodiments and will not be elaborated here. The working condition information can be the specific working condition or the probability of the working condition. What the working condition information specifically is is related to the output data used when building the working condition diagnosis model, and this is not limited in this article.

[0115] In this embodiment, by converting the electrical parameter curve of the pumping unit well into the spectral characteristics of the electrical parameter curve and using the working condition diagnosis model, the working condition can be accurately identified.

[0116] In a further embodiment of this article, in order to further improve the accuracy of working condition identification, as Figure 1B shown, in addition to including the above steps 110 to 130, the intelligent analysis method for the pumping unit well further includes:

[0117] Step 140, calculate the electrical parameter difference curve according to the collected electrical parameter curve and the initial electrical parameter curve of the pumping unit well;

[0118] Step 150, calculate the polished rod load difference curve according to the electrical parameter difference curve;

[0119] Step 160, superimpose the polished rod load difference curve onto the initial indicator diagram to determine the polished rod displacement value;

[0120] Step 170: Draw a dynamometer card curve based on the polished rod displacement value and the polished rod load value obtained after superposition.

[0121] Step 180: Analyze the working condition information of the pumping well by using the dynamometer card curve.

[0122] Step 190: Determine whether the working condition information of the pumping well calculated by using the working condition diagnosis model is consistent with the working condition information of the pumping well analyzed by using the dynamometer card curve. If they are consistent, output the working condition information; if not, send an alarm message.

[0123] In the above Step 140, the initial electrical parameter curve of the pumping well is the electrical parameter curve measured under the initial state of the pumping well. Specifically, when implementing, the following formula can be used to calculate the electrical parameter difference curve:

[0124] ΔP(i) = P1(i) - P0(i);

[0125] where, ΔP(i) is the difference in the input active power of the pumping unit at the i-th moment within a single pumping unit stroke, with the unit of kW; P1(i) is the measured input active power of the pumping unit at the i-th moment within a single pumping unit stroke, with the unit of kW; P0(i) is the initial input active power of the pumping unit at the i-th moment within a single pumping unit stroke, with the unit of kW.

[0126] In the above Step 150, calculating the polished rod load difference curve based on the electrical parameter difference curve includes using the following formula to calculate the polished rod load difference curve:

[0127]

[0128] where, ΔF(i) is the difference in the polished rod load of the pumping well at the i-th moment within a single pumping unit stroke, with the unit of kN; K is the motor efficiency, dimensionless; ΔP(i) is the difference in the input active power of the pumping unit at the i-th moment within a single pumping unit stroke; η1 is the belt drive efficiency; η2 is the speed reducer drive efficiency; i MB is the transmission ratio from the output shaft of the pumping unit to the output shaft of the speed reducer; n m is the motor speed, with the unit of rpm; TF is the torque factor, dimensionless.

[0129] In the above Step 160, the following formula can be used to determine the polished rod displacement value:

[0130] F1(i) = F0(i) + ΔF (i)

[0131] Among them, F1(i) represents the load value corresponding to the i-th point P1(i) within a stroke, with the unit of kN; F0(i) represents the initial load value of the i-th point within a stroke, with the unit of kN, which can be determined from the initial indicator diagram, and ΔF(i) is the polished rod displacement value.

[0132] The implementation processes of the above step 170 and step 180 can refer to the prior art and will not be elaborated here.

[0133] This embodiment can realize the conversion from the electrical parameter curve to the indicator diagram by using the differential pressure method, can eliminate the influence of the balance state of the pumping well, and has the characteristics of simple implementation. In addition, this embodiment can improve the accuracy of working condition identification by comparing the working conditions identified by the two methods and then determining the final working condition.

[0134] In some embodiments, the working condition identified only by the indicator diagram or the working condition identified by the spectral characteristics of the electrical parameter curve can also be used as the final output working condition. In other embodiments, the working condition identified by the spectral characteristics of the electrical parameter curve can also be used as the surface working condition, and the working condition identified by the indicator diagram can be used as the downhole working condition.

[0135] In one embodiment of this article, after obtaining the working condition of the pumping well, in order to effectively guide the optimization of the pumping well, as Figure 2 shown, the intelligent analysis method for pumping wells includes, in addition to the above steps 110 - 130, the following:

[0136] Step 210, according to the working condition of the pumping well, query the optimization database X to obtain the optimization method.

[0137] Among them, the optimization database X stores the corresponding relationship between the working condition and the optimization method, and this relationship can be represented by Table 1 below. Specifically in implementation, this relationship can also be expressed in other ways, and this article does not limit this.

[0138] Table 1:

[0139]

[0140]

[0141] Specifically in implementation, the optimization database Y can also be queried according to the working condition of the pumping well and the type of the pumping well to obtain the optimization method. Among them, the optimization database Y records the corresponding relationship between the working condition type, the type of the pumping well, and the optimization method, and this relationship can be represented by Table 2 below:

[0142] Table 2:

[0143]

[0144] Step 220, control the pumping well according to the obtained optimization method.

[0145] In one embodiment of the present invention, considering the problem that cumulative errors will occur when calculating the production volume using the electrical parameter curve in the prior art (in the prior art, it is necessary to first convert the electrical parameter curve into a dynamometer card and then calculate the production volume using the dynamometer card). As Figure 3 shown, in addition to the above steps 110-step 130, the intelligent analysis method for pumping wells further includes:

[0146] Step 310, collect the production data of the pumping well.

[0147] Specifically, step 310 and step 110 can be executed synchronously. The step numbers in this article are only used to distinguish different steps and are not used to limit the order of steps.

[0148] The production data includes: electrical parameter spectrum, historical change amount, surface stroke frequency, surface stroke, pump diameter, etc.

[0149] Step 320, according to the production data collected in step 310 and the spectrum characteristics of the electrical parameter curve obtained in step 120, calculate the production volume of the pumping well using the production calculation model. That is, input the production data collected in step 310 and the spectrum characteristics of the electrical parameter curve obtained in step 120 into the production calculation model, and the production calculation model will output the production volume of the pumping well.

[0150] The production calculation model is trained by the historical production data corresponding to each production volume of the pumping well and the spectrum characteristics of the historical electrical parameter curve. Specifically, during implementation, first convert the electrical parameter curve into the spectrum characteristics of the electrical parameter curve; then use the historical production data of the pumping well and the spectrum characteristics of the electrical parameter curve as inputs, and the corresponding production volume as the output, and train using a big data model to obtain the production calculation model.

[0151] In some embodiments, considering the following relationship between the production volume and the pump efficiency, the production calculation model can also be composed of a pump efficiency model and the following relational expression. Among them, the pump efficiency model is trained by the historical production data corresponding to each pump efficiency of the pumping well and the spectrum characteristics of the historical electrical parameter curve. The training process refers to the production calculation model and will not be elaborated here.

[0152] Q = A×η×fp×s×n

[0153] Where A is a calculation coefficient, which is related to the units of the stroke, stroke frequency, and pump plunger area. fp is the pump plunger area, which is related to the pump diameter; s is the surface stroke; n is the surface stroke frequency; fp, s, and n are all production data. η is the pump efficiency.

[0154] Furthermore, in order to ensure the accuracy of the optimization of the pumping well, as Figure 4As shown, the intelligent analysis method for pumping wells further includes, in addition to the above steps 110, 310, 120, 320, and 130:

[0155] Step 330: Query the optimization database Z according to the working conditions and production of the pumping well to obtain an optimization method. The optimization database Z stores the corresponding relationships among working conditions, production, and optimization methods, as shown in Table 3:

[0156] Table 3

[0157]

[0158] Step 340: Control the pumping well according to the queried optimization method.

[0159] When this embodiment is implemented, the acquisition of the electrical parameter curve of the pumping well, the diagnosis of the working conditions of the pumping well, the optimization decision-making of the working conditions of the pumping well, and the control based on the Internet of Things can be integrated into an intelligent chip to form a production regulation device for the pumping well, and the closed-loop management of diagnosis-optimization-regulation integration can be realized through this production regulation device for the pumping well.

[0160] In an embodiment of this article, to avoid wasting resources, if a dynamometer card tester is installed on the pumping well, the working conditions of the pumping well are diagnosed according to the collected dynamometer card. For the specific diagnosis method, refer to the prior art and will not be elaborated here. If a dynamometer card tester is not installed on the pumping well, the working conditions of the pumping well are diagnosed using the electrical parameter curve.

[0161] In an embodiment of this article, as Figure 5 shown, the above process of establishing the working condition diagnosis model includes:

[0162] Step 510: Collect the electrical parameter curves of the pumping well under different working conditions.

[0163] Step 530: Perform frequency domain transformation on the collected electrical parameter curves to obtain the frequency spectrum characteristics of each electrical parameter curve.

[0164] Specifically, the frequency spectrum characteristics of the electrical parameter curve are the energy amplitudes of each frequency spectrum, which can be represented in vector form. For example, it can be represented as [d1, d2, … d i , …, dn], where d i is the energy amplitude of the i-th frequency spectrum, and n is the number of frequency spectra. When implemented, considering that among the frequency spectra obtained after the frequency domain transformation of the electrical parameter curve, the energy amplitudes of the first n / 2 and the last n / 2 are symmetric, only s energy amplitudes are taken from the first n / 2 energy amplitudes as features (s <= n / 2) and used as the input of the model in vector form: [d1, d2, …, d s .

[0165] Step 550: Using the working condition or the probability value of the working condition of the pumping well as the output, and the spectral characteristics of the electrical parameter curve under the corresponding working condition as the input, train to obtain a working condition diagnosis model.

[0166] The output of the working condition diagnosis model can be represented in vector form. For example, if there are k types of working conditions, the output of the working condition diagnosis model can be expressed as Y’ = [g1, g2, …, g j …, g k , where any element g j in Y’ is the specific working condition or the probability value of the working condition. When the output of the working condition diagnosis model is represented by the probability value of the working condition, the elements in Y’ satisfy g1 + g2 + … + g k = 1, and the value range of each element is 0 to 1.

[0167] The training process is to gradually correct the parameters in the working condition diagnosis model according to the error of the working condition diagnosis model on the training set until the error reaches the minimum.

[0168] In implementation, considering the characteristic that the sample quantities of the electrical parameter curves of the pumping wells under different working conditions vary greatly, this paper proposes a multi-model integrated modeling method. As Figure 6 shown, step 550 further includes:

[0169] Step 551: Using the working condition or the probability value of the working condition of the pumping well as the output, and the spectral characteristics of the electrical parameter curve under the corresponding working condition as the input, train respectively using a machine learning model and a deep learning model to obtain a first diagnosis model and a second diagnosis model.

[0170] Specifically, the machine learning model includes one of the support vector machine and the K-nearest neighbor method, which has a high recognition rate for working conditions with small sample quantities such as leaking pipes, broken rods, and stuck pumps. The deep learning model includes one of the fully connected network model and the recurrent neural network model, which has a high recognition rate for working conditions with large sample quantities such as normal, insufficient liquid supply, and gas influence. The training processes of the machine learning model and the deep learning model refer to the prior art and will not be elaborated in this paper.

[0171] Step 552: The first diagnosis model and the second diagnosis model constitute the working condition diagnosis model. Specifically, perform weighted processing on the first diagnosis model and the second diagnosis model to obtain a working condition targeting module. In a specific implementation manner, the working condition diagnosis model can be determined through the following formula.

[0172] M = aM1 + bM2;

[0173] Among them, M is the working condition diagnosis model, M1 is the first diagnosis model, M2 is the second diagnosis model, a and b are weights, and the specific values of a and b can be determined according to the proportions of small-sample working conditions and large-sample working conditions. For example, for small-sample working conditions such as pipe leakage, double leakage, rod breakage, and pump sticking, since the sample size is small, the corresponding coefficient a takes a small value; for large-sample working conditions such as normal, insufficient liquid supply, and gas influence, since the sample size is large, the corresponding coefficient b takes a large value. In specific implementation, the weights of a and b can also be set to each account for half, for example, both are 0.5.

[0174] In one embodiment of this article, to ensure the modeling accuracy of the working condition diagnosis model, as Figure 7 shown, after step 510 of collecting the electrical parameter curves of the pumping unit well under different working conditions, the following steps are further included:

[0175] Step 520: Preprocess the electrical parameter curves of the pumping unit well under different working conditions to remove the noise points in the electrical parameter curves, facilitating feature extraction.

[0176] Specifically, the preprocessing process of step 520 includes: respectively performing smoothing, denoising, and standardization processing on the electrical parameter curves of the pumping unit well under different working conditions.

[0177] The electrical parameters of the pumping unit well generally include the current, voltage, active power, and reactive power at the motor input end. Usually, the electrical parameters are divided into three phases: A, B, and C. According to a certain sampling frequency, for the three-phase electrical parameters, in continuous time, n data points of a fixed length are taken, and the same method is used for the preprocessing of each phase of the electrical parameter curve. Specifically, first, if the collected data is in hexadecimal, it needs to be converted to decimal. x is a vector of length n, expanded as: x = [x1, x2,..., x n Each element x i in the vector is a decimal real number. During the collection process, due to hardware problems, some abnormal noise points sometimes occur in the collected data and need to be corrected.

[0178] The processes of smoothing, denoising, and standardization (such as changing the data base) can be referred to in the prior art, and will not be elaborated herein.

[0179] The intelligent analysis method for the pumping unit well provided in this article can accurately determine the working conditions of the pumping unit well based on the electrical parameter curves of the pumping unit well, and can optimize the pumping unit well according to the working condition diagnosis results. Compared with the method of diagnosing working conditions based on the indicator diagram, it can reduce costs, reduce the energy consumption of the oil well, and greatly improve the management level of the oil well.

[0180] Based on the same inventive concept, this article also provides an intelligent analysis device for pumping wells, as described in the following embodiments. Since the principle of solving problems by this device is similar to that of the intelligent analysis method for pumping wells, the implementation of this device can refer to the implementation of the intelligent analysis method for pumping wells, and the repeated parts will not be elaborated. Specifically, as Figure 8A shown, the intelligent analysis device for pumping wells includes:

[0181] A first acquisition module 810, configured to acquire the electrical parameter curve of the pumping well. In specific implementation, the first acquisition module 810 is interconnected with the oilfield Internet of Things of the pumping well, so as to read the electrical parameter data of the pumping well in real time, and obtain the electrical parameter curve according to the electrical parameter data.

[0182] A spectrum transformation module 820, configured to perform frequency-domain transformation on the acquired electrical parameter curve to obtain the spectral characteristics of the electrical parameter curve. The methods of frequency-domain transformation include but are not limited to discrete Fourier transform and wavelet transform.

[0183] A first analysis module 830, configured to calculate the working condition information of the pumping well according to the spectral characteristics of the electrical parameter curve by using a working condition diagnosis model, wherein the working condition diagnosis model is trained by the spectral characteristics of the historical electrical parameter curves under different working conditions of the pumping well.

[0184] Further, as Figure 8B shown, the intelligent analysis device for pumping wells further includes:

[0185] A calculation module 840, configured to calculate an electrical parameter difference curve according to the acquired electrical parameter curve and the initial electrical parameter curve of the pumping well; calculate a polished rod load difference curve according to the electrical parameter difference curve; superimpose the polished rod load difference curve onto the initial indicator diagram to determine the polished rod displacement; draw an indicator diagram curve according to the polished rod displacement value and the polished rod load value obtained after superimposition;

[0186] A second analysis module 850, configured to analyze the working condition information of the pumping well by using the indicator diagram curve;

[0187] A judgment module 860, configured to judge whether the working condition information of the pumping well calculated by using the working condition diagnosis model is consistent with the working condition information of the pumping well analyzed by using the indicator diagram curve. If they are consistent, the working condition information is output; if they are not consistent, an alarm message is sent.

[0188] In some embodiments, please refer back to Figure 5 , the process of establishing the working condition diagnosis model includes:

[0189] Step 510, collecting the electrical parameter curves under different working conditions of the pumping well.

[0190] Step 530: Perform a frequency-domain transformation on each collected electrical parameter curve to obtain the spectral characteristics of each electrical parameter curve.

[0191] Step 550: Use the working condition information of the pumping unit well as the output and the spectral characteristics of the electrical parameter curves under the corresponding working conditions as the input to train a working condition diagnosis model. Specifically, the working condition information can be a specific working condition or a working condition probability value.

[0192] In a further embodiment, considering the characteristic that the sample sizes of the electrical parameter curves of the pumping unit well under different working conditions vary greatly, please refer back to Figure 6 , the above step 550 includes:

[0193] Step 551: Use the working condition or working condition probability value of the pumping unit well as the output and the spectral characteristics of the electrical parameter curves under the corresponding working conditions as the input, and train using a machine learning model and a deep learning model respectively to obtain a first diagnosis model and a second diagnosis model. Specifically, the machine learning model includes one of a support vector machine and the K-nearest neighbor method. The deep learning model includes one of a fully connected network model and a recurrent neural network model.

[0194] Step 552: The working condition diagnosis model is composed of the first diagnosis model and the second diagnosis model. During specific implementation, the working condition diagnosis model can be determined through the following formula.

[0195] M = aM1 + bM2;

[0196] Where M is the working condition diagnosis model, M1 is the first diagnosis model, M2 is the second diagnosis model, a and b are weights, and the specific values of a and b can be determined according to the proportions of small-sample working conditions and large-sample working conditions. For example, if the small-sample working conditions are few, the corresponding coefficient a takes a small value; if the large-sample working conditions are many, the corresponding coefficient b takes a large value. During specific implementation, the weights of a and b can also be set to each account for half, for example, both are 0.5.

[0197] In a further embodiment, to ensure the modeling accuracy of the working condition diagnosis model, please refer back to Figure 7 , after the above step 510 of collecting the electrical parameter curves of the pumping unit well under different working conditions, it further includes:

[0198] Step 520: Preprocess the electrical parameter curves of the pumping unit well under different working conditions.

[0199] Specifically, the preprocessing process of step 520 includes: respectively performing smoothing, denoising, and standardization processing on the electrical parameter curves of the pumping unit well under different working conditions.

[0200] In an embodiment of this article, after obtaining the working conditions of the pumping unit well, to effectively guide the optimization of the pumping unit well, such as Figure 9As shown in the figure, the intelligent analysis device for pumping wells further includes:

[0201] A first optimization module 870, configured to query a first optimization database according to the working conditions of the pumping well to obtain an optimization method, where the first optimization database stores the corresponding relationship between the working conditions and the optimization methods; and control the pumping well according to the queried optimization method.

[0202] In an embodiment of the present invention, considering the problem that cumulative errors will occur when calculating the production volume using the electrical parameter curve in the prior art (in the prior art, it is necessary to first convert the electrical parameter curve into an indicator diagram and then calculate the production volume using the indicator diagram), as Figure 10 As shown in the figure, the intelligent analysis device for pumping wells further includes:

[0203] A second acquisition module 880, configured to acquire the production data of the pumping well.

[0204] A third analysis module 890, configured to calculate the production volume of the pumping well using a production calculation model according to the acquired production data and the spectral characteristics of the electrical parameter curve.

[0205] Wherein, the production calculation model is trained by the historical production data corresponding to each production volume of the pumping well and the spectral characteristics of the historical electrical parameter curve.

[0206] Furthermore, in order to ensure the accuracy of the optimization of the pumping well, as Figure 11 As shown in the figure, the intelligent analysis device for pumping wells further includes:

[0207] A second optimization module 891, configured to query a second optimization database according to the working conditions of the pumping well and the production volume of the pumping well to obtain an optimization method, and control the pumping well according to the queried optimization method. Wherein, the second optimization database stores the corresponding relationship between the working conditions, the production volume and the optimization methods.

[0208] The intelligent analysis device for pumping wells provided in the present invention can eliminate the process of identifying the top and bottom dead points of the electrical parameter curve, eliminate the interference of the balance degree of the pumping well on the characteristics of the electrical parameter curve, and improve the accuracy of working condition diagnosis by using the spectral characteristics of the electrical parameter curve for working condition identification. The conversion of the electrical parameter curve into an indicator diagram can be realized by using the differential pressure method, which can eliminate the influence of the balance state of the pumping well and has the characteristics of simple implementation. In addition, by comparing the working conditions identified by the two methods and then determining the final working condition, the accuracy of working condition identification can be improved.

[0209] To more clearly illustrate the technical solution of the present invention, a specific embodiment is described in detail below. As Figure 12 As shown in the figure, the intelligent analysis method for pumping wells includes:

[0210] Step 1201: Connect to the oilfield Internet of Things, collect the electrical parameter curves of the pumping unit wells from the electrical parameter testers, and at the same time collect the production data of the pumping unit wells. Specifically, the electrical parameter curves of the pumping unit wells include three-phase current, voltage, and active power. The electrical parameter curves should contain at least the information of one complete stroke cycle, and the collection frequency should be high enough, generally not less than 3 Hz.

[0211] Step 1202: Perform preprocessing such as smoothing, denoising, and standardization on the electrical parameter curves of the pumping unit wells collected in Step 1201.

[0212] Step 1203: Perform frequency-domain transformation on the preprocessed electrical parameter curves to obtain the spectral characteristics of the electrical parameter curves: X’ = [d1, d2, …, d i …, d s , d i is one of the spectral characteristics.

[0213] Step 1204: According to the spectral characteristics of the electrical parameter curves, use the working condition diagnosis model to calculate the working condition probability of the pumping unit wells: Y’ = [g1, g2, …, g j …, g k , g j is one of the working condition probabilities. Take the working condition corresponding to the maximum value of the working condition probability as the working condition of the pumping unit well.

[0214] Step 1205: According to the collected production data and the spectral characteristics of the electrical parameter curves, use the production calculation model to calculate the production of the pumping unit wells.

[0215] Step 1206: Convert the preprocessed electrical parameter curves into indicator diagram curves.

[0216] Step 1207: Analyze the working condition information of the pumping unit wells using the indicator diagram curves.

[0217] Step 1208: Determine whether the working condition information of the pumping unit wells calculated using the working condition diagnosis model is consistent with the working condition probability of the pumping unit wells analyzed using the indicator diagram curves. If they are consistent, output the working condition information and execute Step 1209. If they are not consistent, send an alarm message for the staff to determine the working condition.

[0218] Step 1209: According to the working condition of the pumping unit wells obtained in Step 1208 and the production of the pumping unit wells obtained in Step 1205, query the optimization database (as shown in Table 3) to obtain the optimization method.

[0219] Step 1210: Use the optimization method determined in Step 1209 to optimize the pumping unit wells.

[0220] In an embodiment of this article, a computer device is also provided, such as Figure 13As shown, the computer device 1302 may include one or more processors 1304, such as one or more central processing units (CPUs), and each processing unit may implement one or more hardware threads. The computer device 1302 may also include any memory 1306 for storing computer programs that can run on the processor 1304. When the processor 1304 executes the computer program, it implements the intelligent analysis method for pumping wells described in any of the foregoing embodiments. Non-limiting examples include that the memory 1306 may include any one or a combination of the following: any type of RAM, any type of ROM, flash memory devices, hard disks, optical discs, etc. More generally, any memory can store information using any technology. Further, any memory can provide volatile or non-volatile retention of information. Further, any memory can represent a fixed or removable component of the computer device 1302. In one case, when the processor 1304 executes the associated instructions stored in any memory or combination of memories, the computer device 1302 can perform any operation of the associated instructions. The computer device 1302 also includes one or more drive mechanisms 1308 for interacting with any memory, such as a hard disk drive mechanism, an optical disc drive mechanism, etc.

[0221] The computer device 1302 may also include an input / output module 1310 (I / O) for receiving various inputs (via the input device 1312) and for providing various outputs (via the output device 1314). A specific output mechanism may include a presentation device 1316 and an associated graphical user interface 1318 (GUI). In other embodiments, the input / output module 1310 (I / O), the input device 1312, and the output device 1314 may not be included, and it may only be a computer device in a network. The computer device 1302 may also include one or more network interfaces 1320 for exchanging data with other devices via one or more communication links 1322. One or more communication buses 1324 couple the components described above together.

[0222] The communication link 1322 can be implemented in any way, for example, through a local area network, a wide area network (e.g., the Internet), a point-to-point connection, etc., or any combination thereof. The communication link 1322 may include any combination of hardwired links, wireless links, routers, gateway functions, name servers, etc. governed by any protocol or combination of protocols.

[0223] In one embodiment herein, a computer-readable storage medium is also provided. The computer-readable storage medium stores a computer program, and when the computer program is run by a processor, it executes the steps of the intelligent analysis method for pumping wells described in any of the above examples.

[0224] An embodiment of this article also provides a computer-readable instruction. When a processor executes the instruction, the program therein causes the processor to execute the steps of the intelligent analysis method for a pumping unit well described in any of the above examples.

[0225] It should be understood that in various embodiments of this article, the magnitudes of the serial numbers of the above processes do not mean the sequence of execution. The execution sequence of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of this article.

[0226] It should also be understood that in the embodiments of this article, the term "and / or" is only a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent three situations: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this article generally represents an "or" relationship between the associated objects before and after.

[0227] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed in this article can be implemented by electronic hardware, computer software, or a combination of the two. To clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described according to functions in the above description. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this article.

[0228] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0229] In several embodiments provided in this article, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed coupling or direct coupling or communication connection to each other can be an indirect coupling or communication connection through some interfaces, devices, or units, and can also be in an electrical, mechanical, or other form of connection.

[0230] The unit described as a separation component may or may not be physically separated. The component shown as a unit may or may not be a physical unit, that is, it may be located in one place or distributed over multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiments in this article.

[0231] In addition, each functional unit in the various embodiments in this article may be integrated into a processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

[0232] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the essence of the technical solution in this article, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to enable a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this article. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0233] Specific embodiments are used in this article to elaborate on the principles and implementation manners of this article. The description of the above embodiments is only used to help understand the method and its core idea in this article; at the same time, for those of ordinary skill in the art, according to the idea in this article, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to this article.

Claims

1. An intelligent analysis method for pumping wells, characterized in that, Including: Collecting the electrical parameter curves of the pumping unit well; Performing frequency-domain transformation on the collected electrical parameter curves to obtain the spectral characteristics of the electrical parameter curves; According to the spectral characteristics of the electrical parameter curves, using a working condition diagnosis model to calculate the working condition information of the pumping unit well; wherein, the working condition diagnosis model is trained by the spectral characteristics of the historical electrical parameter curves under different working conditions of the pumping unit well; Calculating an electrical parameter difference curve according to the collected electrical parameter curves and the initial electrical parameter curves of the pumping unit well; Calculating a polished rod load difference curve according to the electrical parameter difference curve; Superimposing the polished rod load difference curve onto the initial indicator diagram to determine the polished rod displacement value; Drawing an indicator diagram curve according to the polished rod displacement value and the polished rod load value obtained after superposition; Analyzing the working condition information of the pumping unit well by using the indicator diagram curve; Judging whether the working condition information of the pumping unit well calculated by using the working condition diagnosis model is consistent with the working condition information of the pumping unit well analyzed by using the indicator diagram curve. If they are consistent, output the working condition information; if they are not consistent, send an alarm message.

2. The method according to claim 1, wherein Calculating the polished rod load difference curve according to the electrical parameter difference curve, including calculating the polished rod load difference curve by using the following formula: Among them, DF(i) is the difference in polished rod load of the pumping well at the i-th moment within a single stroke of the pumping unit, with the unit of kN; K is the motor efficiency, dimensionless; DP(i) is the difference in active power input to the pumping unit at the i-th moment within a single stroke of the pumping unit; η1 is the belt drive efficiency; η2 is the speed reducer drive efficiency; i MB is the transmission ratio from the output shaft of the pumping unit to the output shaft of the speed reducer; n m is the motor speed, with the unit of rpm; TF is the torque factor, dimensionless.

3. The method according to claim 1, wherein Also including: Querying a first optimization database according to the working condition of the pumping unit well to obtain an optimization method, wherein the first optimization database stores the corresponding relationship between the working condition and the optimization method; Controlling the pumping unit well according to the queried optimization method.

4. The method according to claim 1, characterized in that Also including: Collecting the production data of the pumping unit well; Calculating the production of the pumping unit well by using a production calculation model according to the collected production data and the spectral characteristics of the electrical parameter curves; Wherein, the production calculation model is trained by the historical production data corresponding to each production of the pumping unit well and the spectral characteristics of the historical electrical parameter curves.

5. The method according to claim 4, wherein Also including: Querying a second optimization database according to the working condition of the pumping unit well and the production of the pumping unit well to obtain an optimization method, wherein the second optimization database stores the corresponding relationship between the working condition, the production and the optimization method; Controlling the pumping unit well according to the queried optimization method.

6. The method according to claim 1, wherein The process of establishing the working condition diagnosis model includes: Collecting the electrical parameter curves of the pumping unit well under different working conditions; Performing frequency-domain transformation on each collected electrical parameter curve to obtain the spectral characteristics of each electrical parameter curve; Using the working condition information of the pumping unit well as the output and the spectral characteristics of the electrical parameter curves under the corresponding working conditions as the input to train the working condition diagnosis model.

7. The method according to claim 6, wherein Using the working condition information of the pumping unit well as the output and the spectral characteristics of the electrical parameter curves under the corresponding working conditions as the input to train the working condition diagnosis model, including: Using the working condition or the working condition probability value of the pumping unit well as the output and the spectral characteristics of the electrical parameter curves under the corresponding working conditions as the input, and respectively training by using a machine learning model and a deep learning model to obtain a first diagnosis model and a second diagnosis model; The working condition diagnosis model is composed of the first diagnosis model and the second diagnosis model.

8. The method according to claim 7, wherein The working condition diagnosis model is composed of the first diagnosis model and the second diagnosis model, including: Perform a weighted process on the first diagnostic model and the second diagnostic model to obtain the working condition diagnostic model.

9. The method according to claim 7, wherein The machine learning model includes one of a support vector machine and the k-nearest neighbor method; the deep learning model includes one of a fully connected network model and a recurrent neural network model.

10. The method according to claim 6, characterized in that, After collecting the electrical parameter curves of the pumping well under different working conditions, it further includes: Perform preprocessing on the electrical parameter curves of the pumping well under different working conditions.

11. The method according to claim 10, wherein Performing preprocessing on the electrical parameter curves of the pumping well under different working conditions includes: Perform smoothing, denoising, and standardization processing on the electrical parameter curves of the pumping well under different working conditions respectively.

12. An intelligent analysis device for pumping wells, characterized in that, It includes: A first acquisition module for acquiring the electrical parameter curves of the pumping well; A frequency spectrum transformation module for performing a frequency domain transformation on the acquired electrical parameter curves to obtain the frequency spectrum characteristics of the electrical parameter curves; A first analysis module for calculating the working condition information of the pumping well by using the working condition diagnostic model according to the frequency spectrum characteristics of the electrical parameter curves; wherein, the working condition diagnostic model is trained by the frequency spectrum characteristics of the historical electrical parameter curves of the pumping well under different working conditions; A calculation module for calculating an electrical parameter difference curve according to the acquired electrical parameter curves and the initial electrical parameter curves of the pumping well; calculating a polished rod load difference curve according to the electrical parameter difference curve; superimposing the polished rod load difference curve onto the initial indicator diagram to determine the polished rod displacement value; drawing an indicator diagram curve according to the polished rod displacement value and the polished rod load value obtained after superimposition; A second analysis module for analyzing the working condition information of the pumping well by using the indicator diagram curve; A judgment module for judging whether the working condition information of the pumping well calculated by using the working condition diagnostic model is consistent with the working condition information of the pumping well analyzed by using the indicator diagram curve. If they are consistent, the working condition information is output. If they are not consistent, an alarm message is sent.

13. A computer device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the intelligent analysis method for the pumping well according to any one of claims 1 to 11.

14. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, and when the computer program is executed by the processor, it implements the intelligent analysis method for the pumping well according to any one of claims 1 to 11.

Images