CN109447380B

CN109447380B - Method and device for determining oil well yield

Info

Publication number: CN109447380B
Application number: CN201810643402.1A
Authority: CN
Inventors: 葛婷; 李志元; 任卓
Original assignee: Beijing Gridsum Technology Co Ltd
Current assignee: Beijing Gridsum Technology Co Ltd
Priority date: 2018-06-21
Filing date: 2018-06-21
Publication date: 2021-01-22
Anticipated expiration: 2038-06-21
Also published as: CN109447380A

Abstract

The invention discloses a method and a device for determining oil well yield, which can obtain parameter values of various working parameters of a first oil well through a sensor and an indicator diagram respectively, and input the parameter values into an oil well yield model of the first oil well to obtain the oil well yield of the first oil well. Because the oil well production model is a model of a single oil well and is obtained by machine learning according to the training data of the oil well, the oil well production model used by the invention has stronger pertinence to the single oil well. Therefore, the accuracy of the oil well yield determined by the method is higher. Meanwhile, the oil well yield is obtained through the oil well yield model obtained through machine learning, so that the method does not need to perform complicated operation on the oil well yield, and is more convenient and quicker.

Description

Method and device for determining oil well yield

Technical Field

The invention relates to the field of fossil fuel production, in particular to a method and a device for determining oil well yield.

Background

Fossil fuels are an indispensable energy source in the modern society, and the determination of oil well production is very important in the field of fossil fuel production.

In the prior art, various parameters such as oil-water mixture density, gas-oil ratio, sucker rod weight, friction coefficient, indicator diagram parameters and the like are collected, and then complex operation is carried out according to a plurality of formulas to obtain the oil well yield. The prior art applies the above formula to the well production calculations for all wells. However, the inventor of the present application has found that the production of the oil wells determined by the formula of the prior art is different from one oil well to another, and the production of the oil wells determined by the formula of the prior art is larger in error for some oil wells.

It can be seen that the accuracy of well productivity determined by existing well productivity determination schemes is low.

Disclosure of Invention

In view of the above, the present invention has been made to provide a method and apparatus for determining well production that overcomes or at least partially solves the above problems, by:

a method of determining well production, comprising:

obtaining parameter values of a plurality of operating parameters of the first well over a first time period by a plurality of sensors disposed on the production equipment of the first well;

obtaining parameter values of various working parameters of the first oil well in the first time period according to an indicator diagram of oil extraction equipment of the first oil well, wherein the types of the working parameters obtained through the indicator diagram are different from the types of the working parameters obtained through the sensor;

and inputting the obtained parameter values into a well production model of the first well to obtain the well production of the first well output by the well production model, wherein the well production model of the first well is a model for outputting the well production of the first well obtained by machine learning of the training data of the first well.

Optionally, the training data of the first well comprises: the production of the first oil well in a historical time period, the parameter values of the various operating parameters of the first oil well in the historical time period obtained by various sensors arranged on the oil production equipment of the first oil well, and the parameter values of the various operating parameters of the first oil well in the historical time period obtained according to an indicator diagram of the oil production equipment of the first oil well;

alternatively, the training data for the first well comprises: the method comprises the steps of obtaining well production of at least one other well in the same field development block as the first well in a historical time period, obtaining parameter values of various operating parameters of the at least one other well in the historical time period by various sensors arranged on oil production equipment of the at least one other well, and obtaining parameter values of various operating parameters of the at least one other well in the historical time period according to indicator diagrams of the oil production equipment of the at least one other well.

Optionally, the oil extraction device is an oil pumping unit, and the plurality of working parameters obtained by the plurality of sensors include: at least one of indicator diagram load, pumping unit stroke, oil pressure, casing pressure, working current of the motor of the pumping unit and working voltage of the motor of the pumping unit,

and/or the oil extraction equipment is an oil pumping unit, and the multiple working parameters obtained according to the indicator diagram comprise: and at least one of the effective stroke, the area of the indicator diagram and the stroke frequency of the oil extraction machine.

Optionally, the method further includes:

obtaining an actual well production for the first well over the first time period;

and determining the error of the oil well production output by the oil well production model of the first oil well relative to the actual oil well production, and performing machine learning on each obtained parameter value and the actual oil well production when the error is larger than a threshold value so as to adjust the oil well production model of the first oil well.

Optionally, the machine learning is performed on the training data to obtain: a weight of an effect of each of the operating parameters on well production from the first well, the method further comprising:

and outputting the influence weight.

Optionally, the inputting the obtained parameter values into the well production model of the first well to obtain the well production of the first well output by the well production model includes:

and inputting the obtained parameter values into the oil well production model of the first oil well, and obtaining the oil well production of the first oil well, which is determined and output by the oil well production model according to the influence weight and the input parameter values.

An oil well production determination apparatus comprising: a first parameter value obtaining unit, a second parameter value obtaining unit, and a first yield obtaining unit,

the first parameter value obtaining unit is used for obtaining parameter values of various working parameters of the first oil well in a first time period through various sensors arranged on oil production equipment of the first oil well;

the second parameter value obtaining unit is configured to obtain parameter values of multiple working parameters of the first oil well in the first time period according to an indicator diagram of the oil production equipment of the first oil well, where the types of the working parameters obtained through the indicator diagram are different from the types of the working parameters obtained through the sensor;

and the first yield obtaining unit is used for inputting the obtained parameter values into a well yield model of the first well to obtain the well yield of the first well output by the well yield model, wherein the well yield model of the first well is a model which is obtained by machine learning of the training data of the first well and is used for outputting the well yield of the first well.

A storage medium comprising a stored program, wherein the program, when executed, controls a device on which the storage medium is located to perform the method of determining well production as described above.

A processor for executing a program, wherein the program when executed performs the method of well production determination described above.

By means of the technical scheme, the method and the device for determining the oil well yield can obtain the parameter values of various working parameters of the first oil well through the sensors and the indicator diagram respectively, and input the parameter values into the oil well yield model of the first oil well to obtain the oil well yield of the first oil well. Because the oil well production model is a model of a single oil well and is obtained by machine learning according to the training data of the oil well, the oil well production model used by the invention has stronger pertinence to the single oil well. Therefore, the accuracy of the oil well yield determined by the method is higher. Meanwhile, the oil well yield is obtained through the oil well yield model obtained through machine learning, so that the method does not need to perform complicated operation on the oil well yield, and is more convenient and quicker.

The foregoing description is only an overview of the technical solutions of the present invention, and the embodiments of the present invention are described below in order to make the technical means of the present invention more clearly understood and to make the above and other objects, features, and advantages of the present invention more clearly understandable.

Drawings

Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the invention. Also, like reference numerals are used to refer to like parts throughout the drawings. In the drawings:

FIG. 1 is a flow chart of a method for determining well production according to an embodiment of the present invention;

FIG. 2 is a schematic diagram of an indicator diagram provided by an embodiment of the invention;

FIG. 3 is a flow chart of another method for determining well production provided by embodiments of the present invention;

fig. 4 is a schematic structural diagram of an oil well production determining apparatus according to an embodiment of the present invention.

Detailed Description

Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

As shown in fig. 1, an embodiment of the present invention provides a method for determining oil well production, which may include:

s100, obtaining parameter values of various working parameters of the first oil well in a first time period through various sensors arranged on oil extraction equipment of the first oil well;

wherein, oil recovery equipment can be beam-pumping unit, through the multiple working parameter that multiple sensor obtained can include: at least one of indicator diagram load, pumping unit stroke, oil pressure, casing pressure, operating current of a motor of the pumping unit and operating voltage of the motor of the pumping unit.

The unit is kilonewton, and the diameter change caused by the elongation change of the sucker rod by the indicator diagram sensor can be measured and calculated. Specifically, the indicator diagram sensor has various forms, such as: the embodiment of the present invention is not limited herein. Besides the indicator diagram load, the load and displacement integrated sensor can also measure the moving distance, namely the displacement, of the sucker rod, so that an indicator diagram consisting of 144 displacement and corresponding data points of the load is generated. The indicator diagram can be shown in fig. 2, and is a graph reflecting the working condition of the pumping unit, and generally, the abscissa is displacement and the ordinate is load.

Specifically, the pumping unit stroke is the maximum displacement of a pump piston (the distance between the highest point and the lowest point of the pump piston) of the pumping unit, and the pumping unit stroke can be acquired by an indicator diagram sensor. According to the embodiment of the invention, the displacement corresponding to the 73 th point in the indicator diagram can be used as the stroke of the oil extraction machine.

The oil pressure is the annular pressure of an oil pipe measured by a pressure gauge during production of the oil production well, and the casing pressure is the annular pressure between the oil pipe and the casing measured by the pressure gauge during production of the oil production well.

The working current and the working voltage of the motor of the oil extraction machine can be measured by the ammeter and the voltmeter respectively, and certainly, in practical application, when the working voltage of the motor of the oil extraction machine is the rated voltage, the rated voltage of the motor can be directly determined as the working voltage of the motor by the method and the device, and the voltmeter is not required to be used for measurement.

Alternatively, the first time period may be one production cycle of well production or a time period of any duration set by a technician, such as a day.

In practical application, the duration of the first time period may be twenty minutes, and the present invention may collect parameter values of a plurality of working parameters for a plurality of times according to the duration of the first time period as a cycle.

Due to the limitation of the formula adopted by the prior art, any one of the working parameters of oil pressure, casing pressure, the working current of the motor of the oil production machine and the working voltage of the motor of the oil production machine is not taken as the working parameter required for calculating the yield of the oil well in the prior art. The inventor of the application finds that: the oil pressure, the casing pressure, the operating current of the motor of the oil extraction machine and the operating voltage of the motor of the oil extraction machine also have a certain influence on the oil well production, for example: for the same oil well, when the working current of the motor of the oil extraction machine and the working voltage of the motor of the oil extraction machine are higher, the oil well yield of the oil well is also higher. Therefore, the invention breaks through the limitation of the formula in the prior art, and creatively takes at least one of the indicator diagram load, the stroke, the oil pressure, the casing pressure, the working current of the motor of the oil extraction machine and the working voltage of the motor of the oil extraction machine as the input of the oil well yield model for calculating the oil well yield. Thus, the invention considers more working parameters which can influence the oil well yield, and can effectively improve the accuracy of the determined oil well yield.

S200, obtaining parameter values of various working parameters of the first oil well in the first time period according to an indicator diagram of oil extraction equipment of the first oil well;

the types of the working parameters obtained through the indicator diagram may be different from the types of the working parameters obtained through the sensor. Specifically, the plurality of operating parameters obtained according to the indicator diagram may include: and at least one of the effective stroke, the area of the indicator diagram and the stroke frequency of the oil extraction machine. It can be understood that the various operating parameters obtained through the indicator diagram can be obtained only by analyzing and calculating the indicator diagram.

Specifically, the effective stroke of the pumping unit is the effective distance that the oil well pump piston of the pumping unit actually drives the liquid to do work in the up-and-down motion process in the pump cylinder, and the effective distance comprises: an up active stroke and a down active stroke. The effective stroke of the oil extraction machine can be obtained by the following steps: the method comprises the steps of firstly carrying out low-pass filtering smoothing treatment on 144 corresponding data points of displacement and load in an indicator diagram, then obtaining up-down stroke inflection points of the 144 data points of the indicator diagram through curvature derivation, and dividing a ring-shaped curve into single-value curves to obtain the effective stroke of the oil extraction machine, namely the effective value of the up-down stroke.

Wherein, the frequency of stroke is the frequency of the up-and-down movement of the piston of the oil pump of the oil extraction machine in the working cylinder per minute. The measuring period of a complete indicator diagram (the measuring time of a complete indicator diagram) directly reflects the movement period of the oil well pump piston, and the number of the measuring periods contained in one minute time is the number of times of stroke.

The indicator diagram area is an area of an irregular shape formed by 144 points of the indicator diagram, and the indicator diagram area can be obtained by the following steps: the method comprises the steps of cutting an irregular shape formed by 144 points of the indicator diagram into regular shapes (such as triangles or quadrangles), calculating the area of each regular shape respectively, and accumulating to obtain the area of the indicator diagram.

The indicator diagram is an important basis for knowing the working condition of oil production equipment in oil and gas production. Many abnormal phenomena in the working process of the oil extraction equipment can be reflected on the indicator diagram more visually, and reasonable well switching time can be worked out for low-yield and low-energy wells through the indicator diagram method, so that the abrasion of the oil extraction equipment, the waste of electric energy and the like are reduced. Similarly, there is a very close relationship between the various operating parameters analyzed and calculated from the indicator diagram and the well production, and when the operating parameters in the second parameter set change, these changes will affect the well production to a greater extent and in a shorter time. Therefore, it is important to take the various operating parameters obtained from the indicator diagram as inputs to the well production model.

In the prior art, working parameters such as oil-water mixture density, gas-oil ratio, sucker rod weight, friction coefficient and the like are usually used as important parameters influencing the yield of an oil well for collection and use. However, the working parameters selected by the embodiment of the invention may not include working parameters such as oil-water mixing density, gas-oil ratio, sucker rod weight, friction coefficient and the like. This is because the inventor of the present application finds out in the process of implementing the present invention that: the frequency of changes in oil-water mixture density and gas-oil ratio is high, but these parameters can only be collected at longer intervals. In order to correspond to other working parameters which can be acquired in real time, the prior art needs to estimate a plurality of oil-water mixing densities and a plurality of gas-oil ratios in the time interval. This results in less accurate calculated well production because the estimated values of the operating parameters are inaccurate. Therefore, the invention can effectively improve the accuracy of oil well yield by removing the oil-water mixed density and the gas-oil ratio. Meanwhile, for the same oil well, the values of the weight and the friction coefficient of the sucker rod are generally fixed and unchangeable numbers, and the existence of the fixed and unchangeable numbers in training data in machine learning has no influence on a training result, so the invention can eliminate the weight and the friction coefficient of the sucker rod in order to simplify the machine learning process and the oil well yield determination process.

Meanwhile, the traditional mode of calculating the oil well yield by using a formula only selects indicator diagram data to participate in calculation, and the oil pressure, the casing pressure, the working current of the motor of the oil extraction machine and the working voltage of the motor of the oil extraction machine are added as working parameters influencing the oil well yield, so that the accuracy of the oil well yield output by the oil well yield model is improved.

And S300, inputting the obtained parameter values into a well yield model of the first well to obtain the well yield of the first well output by the well yield model, wherein the well yield model of the first well is a model for outputting the well yield of the first well obtained by machine learning of the training data of the first well.

In particular, the training data for the first well may be historical production data for the first well or historical production data for at least one other well located in the same field development block as the first well.

In practical application, the method can divide the obtained historical production data into two parts, wherein one part is used as training data, and the other part is used as test data. For example: the invention obtains 1 ten thousand historical production data, each historical production data comprises parameter values of a plurality of production parameters corresponding to one moment. 8000 parts of historical production data are used as training data, and the other 2000 parts of historical production data are used as test data. Firstly, machine learning is carried out through training data to obtain an oil well production model, then accuracy testing is carried out on the oil well production model through testing data, and if the testing fails, the testing data is added or modified to continue machine learning until the testing passes.

The historical production data of the oil well may include production data under a plurality of different working conditions, for example: normal production, pumping and spraying, serious leakage and the like. The production data under different working conditions are used as training data together, so that the oil well yield model obtained after machine learning can output the matched oil well yield according to the input production data under different working conditions.

During operation of oil recovery equipment in an oil well, when formation pressure suddenly increases, fluid in the formation may overcome the pressure of the fluid column in the entire wellbore and flow directly out of the surface, a phenomenon known as pump-out. Specifically, the invention can determine whether the production data is in the pumping state according to the oil pressure and the downlink load. For example: comparing the oil pressure at two moments separated by a certain time interval, comparing the downlink load at the two moments, and when the oil pressure and the downlink load are determined to be increased more, determining that the blowout occurs in the time period of the two moments, and determining the production data in the time period as the production data in the blowout state.

Correspondingly, the invention can determine whether the serious leakage occurs according to the upper effective stroke and the lower effective stroke, specifically, the invention can compare the upper effective strokes at two moments separated by a certain time interval and the lower effective strokes at the two moments, when the upper effective stroke and/or the lower effective stroke are obviously reduced, the time period of the two moments is determined to have the serious leakage, and the production data in the time period is determined as the production data in the serious leakage state.

After a first well has been provided with various sensors on the production facility and has been producing for a period of time, embodiments of the present invention may obtain historical production data for the first well over the historical period of time, which may include, in particular: the production of the first oil well in the historical time period, the parameter values of the various working parameters of the first oil well in the historical time period obtained by various sensors arranged on the oil production equipment of the first oil well, and the parameter values of the various working parameters of the first oil well in the historical time period obtained according to the indicator diagram of the oil production equipment of the first oil well. Namely: the training data for the first well may include: the method comprises the steps of obtaining the production of the first oil well in a historical time period, obtaining parameter values of various working parameters of the first oil well in the historical time period by various sensors arranged on a production device of the first oil well, and obtaining the parameter values of the various working parameters of the first oil well in the historical time period according to an indicator diagram of the production device of the first oil well.

When the first well is a new well or the first well has just provided various sensors on the production equipment, the historical production data of the first well will not exist or be less, and at this time, the historical production data of at least one other well located in the same field development zone as the first well can be used as the training data of the first well. Namely: the training data for the first well may include: the method comprises the steps of obtaining well production of at least one other well in the same field development block as the first well in a historical time period, obtaining parameter values of various operating parameters of the at least one other well in the historical time period by various sensors arranged on oil production equipment of the at least one other well, and obtaining parameter values of various operating parameters of the at least one other well in the historical time period according to indicator diagrams of the oil production equipment of the at least one other well. It will be appreciated that there is a high similarity between different wells located in the same field development block, and therefore a well production model for a first well can be obtained from historical production data for other wells within the same field development block.

Specifically, the training data of the first well may be as shown in table 1 (in table 1, "operating current of the motor of the oil production machine" is represented by "operating current", and "operating voltage of the motor of the oil production machine" is represented by "operating voltage"), and the "oil production machine active stroke" includes an upper active stroke and a lower active stroke, and the indicator diagram load is large, and only the data of the indicator diagram load 89 is exemplarily provided):

TABLE 1

The unit of the indicator diagram load is kilonewton, the unit of the oil pressure and the casing pressure is megapascal, the unit of the working current is ampere, the unit of the working voltage is volt, the unit of the upper effective stroke, the unit of the lower effective stroke and the unit of the oil extraction machine stroke are all millimeters, and the area of the indicator diagram is a relative value and is dimensionless. The unit of well production is ton.

Specifically, the input of the well production model of the first well may be parameter values of the plurality of operating parameters obtained in steps S100 and S200, and the output of the well production model of the first well may be the well production of the first well. The invention can also obtain the influence weight of each working parameter on the oil well yield of the first oil well after the machine learning is carried out on the training data. The influence weight of a certain working parameter is a quantitative index of the oil well yield change caused by the working parameter change, and when the influence weight of the working parameter is larger, the oil well yield change caused by the working parameter change is larger, and vice versa, the oil well yield change is smaller.

For example: for the training data shown in table 1, the present invention can obtain the impact weights shown in table 2:

TABLE 2

Among them, there are various machine learning methods, such as: k-nearest neighbors, logistic regression, gradient boosting trees, support vector machines, naive bayes, decision trees, random forests, and the like. In practical application, the embodiment of the invention can select one of the machine learning methods to perform machine learning on the training data in sequence so as to obtain the oil well yield models corresponding to different machine learning methods. Then, the embodiment of the invention can carry out error detection on each oil well yield model and take the oil well yield model with the minimum error as the preset oil well yield model.

In practical application, the invention can respectively use the production data of each oil well as the training data, thereby respectively establishing the oil well production model aiming at each oil well according to the training data. Therefore, the method can establish the oil well yield model for each oil well in different geographical positions and wax precipitation conditions, so that the accuracy of the oil well yield model is higher. In practical application, for any oil well, the invention can also respectively use the production data of the oil well in different periods (such as a certain geological period and a certain production period) or different production conditions (such as pumping and severe loss) as the training data, and establish the oil well production model aiming at the different periods or different production conditions of the oil well according to the training data of the oil well in different periods or different production conditions. Therefore, the oil well yield model obtained by the method is more refined, and the accuracy of the oil well yield output by the oil well yield model can be further improved.

Specifically, after obtaining a plurality of well production models of the same well, the present invention can store model matching conditions corresponding to the respective well production models of the well. The model matching conditions are used to determine a well production model that matches the obtained parameter values of the operating parameters of the well. The model matching condition may be a range of values of one or more operating parameters. The model matching condition can be set and modified by workers according to actual conditions.

After the parameter values of various working parameters are obtained, the oil well yield model matched with the obtained parameter values of the working parameters can be determined according to model matching conditions; the obtained parameter values are then input into the determined well production model. For example: the model matching conditions corresponding to the oil well yield model under severe loss are as follows: the active stroke is within a first range of values. The present invention can determine a well production model under severe loss as a well production model matching the parameter values of the obtained operating parameters when the effective stroke among the plurality of operating parameters obtained by the present invention is within the first range of values. Therefore, the method also realizes automatic identification of the oil well yield model matched with the obtained parameter values, does not need manual identification, and has high accuracy.

According to the method for determining the oil well yield, the parameter values of the various working parameters of the first oil well can be obtained through the sensors and the indicator diagram respectively, and the parameter values are input into the oil well yield model of the first oil well to obtain the oil well yield of the first oil well. Because the oil well production model is a model of a single oil well and is obtained by machine learning according to the training data of the oil well, the oil well production model used by the invention has stronger pertinence to the single oil well. Therefore, the accuracy of the oil well yield determined by the method is higher. Meanwhile, the oil well yield is obtained through the oil well yield model obtained through machine learning, so that the method does not need to perform complicated operation on the oil well yield, and is more convenient and quicker.

Specifically, when the oil well yield model is established, the embodiment of the invention can perform butt joint modeling on each working parameter in the training data and the oil well yield, so as to determine the matching relationship between the working parameter and the oil well yield. Specifically, the matching relationship may be embodied in a manner of affecting the weight. Namely: the oil well yield model can also obtain the following results after machine learning is carried out on training data: a weight of an effect of each of the operating parameters on well production from the first well. On this basis, the method shown in fig. 1 may further include:

the impact weight is output.

The step S300 of the method shown in fig. 1 of inputting the obtained parameter values into the well production model of the first well to obtain the well production of the first well output by the well production model may specifically include:

Wherein, the influence weight can be determined according to whether the oil well yield changes along with the change of the working parameters and the change amplitude of the oil well yield along with the change of the working parameters. When the oil well production does not change no matter how a certain working parameter changes, the working parameter is not influenced by the oil well production (the working parameter can also be described as being influenced by the oil well production by 0). The invention can respectively determine the variation amplitude of the oil well yield caused by the variation of different working parameters in the same amplitude, thereby determining the size of the influence weight according to the variation amplitude of the oil well yield. Specifically, the greater the magnitude of the change in well production, the greater the weight of the impact.

By outputting the influence weight, the embodiment of the invention can enable oil field workers to obtain the influence of each working parameter on the oil well yield, so that each working parameter is adjusted according to the influence weight to achieve the maximization of the oil well yield or the maximization of benefit.

Specifically, the well production output by the well production model may be the well production in a first time period, or may be the well production in a second time period including the first time period. In practical applications, the oil well production model in the embodiment of the present invention may first obtain the oil well production in the first time period, and then obtain the oil well production in the second time period according to the time relationship between the second time period and the first time period. For example: the duration of the first period of time is twenty minutes and the duration of the second period of time may be 24 hours, i.e. 1 day. The present invention may first obtain the well production for the first time period and then multiply that well production by 72 to obtain the result of the calculation as the well production for the second time period. Specifically, the second time period may be one production cycle. Therefore, the yield of the oil extraction equipment continuously working for one production cycle under the current working parameters can be obtained, and the analysis and comparison of the yield by oil field workers are facilitated.

As shown in fig. 3, another method for determining oil well production according to the embodiment of the present invention may further include:

s500, obtaining the actual oil well yield of the first oil well in the first time period;

s600, determining an error of the oil well yield output by the oil well yield model of the first oil well relative to the actual oil well yield, and when the error is larger than a threshold value, performing machine learning on each obtained parameter value and the actual oil well yield to adjust the oil well yield model of the first oil well.

The actual oil well yield can be input by an oil field worker or sensed by a yield metering device.

The error of the oil well production relative to the actual oil well production may be a difference between the oil well production output by the oil well production model and the actual oil well production, a percentage of the difference relative to the actual oil well production or the oil well production output by the oil well production model, or other forms, and the embodiment of the present invention is not limited herein.

Of course, in practical applications, the technician may modify or add the training data of the well production model periodically or at any time to improve the accuracy of the well production output by the well production model in time. For example: every other month, various operating parameters of the first oil well obtained within the last month and the oil well production of the first oil well within the last month are added to the training data and the training data are machine-learned again. Thus, the present invention achieves iterative optimization of the well production model.

The modification or addition of the training data does not result in a change to the machine learning algorithm of the well production model, nor does the present invention require modification of the machine learning algorithm of the well production model. Since the modification and addition of training data is very simple for the technician, the technician does not need to know the machine learning algorithm, greatly reducing the requirements on the technician. For example: the device for implementing the oil well production determining method provided by the embodiment of the invention is composed of a client device and a server which are connected in a communication mode, wherein the client device executes steps S100 and S200 shown in FIG. 1 and sends obtained parameter values to the server. The server performs step S300 to obtain the well production. Further, the server may also return the resulting well production to the client device, which may output it to inform the client. Therefore, in the process, the user only needs to input all the working parameters into the client device, and the method is simple and convenient. Meanwhile, in the training process of the oil well yield model, the client device sends the training data to the server, and the server performs machine learning on the training data. When a user wants to optimize the oil well yield model, the user only needs to input each new training data into the client device, and does not need to know the specific machine learning process, so that the operation burden of the user is greatly reduced.

The embodiment of the invention shown in fig. 3 can use the parameter values of the working parameters and the actual oil well yield in the first time period as training data when the error of the oil well yield output by the oil well yield model relative to the actual oil well yield is larger than the threshold value, and control the oil well yield model to perform machine learning on the training data. Through the machine learning, the oil well yield model can be optimized, and therefore the accuracy of the output oil well yield is improved.

Corresponding to the oil well yield determination method, the embodiment of the invention also provides an oil well yield determination device.

As shown in fig. 4, an oil well production determining apparatus provided by an embodiment of the present invention may include: a first parameter value obtaining unit 100, a second parameter value obtaining unit 200 and a first yield obtaining unit 300,

the first parameter value obtaining unit 100 is used for obtaining parameter values of various working parameters of the first oil well in a first time period through various sensors arranged on oil production equipment of the first oil well;

The second parameter value obtaining unit 200 is configured to obtain parameter values of multiple working parameters of the first oil well in the first time period according to an indicator diagram of oil recovery equipment of the first oil well, where the types of the working parameters obtained through the indicator diagram are different from the types of the working parameters obtained through the sensor;

The first production obtaining unit 300 is configured to input each obtained parameter value into a well production model of the first well, and obtain the well production of the first well output by the well production model, where the well production model of the first well is a model for outputting the well production of the first well obtained by machine learning of the training data of the first well.

In practical application, the method can divide the obtained historical production data into two parts, wherein one part is used as training data, and the other part is used as test data.

Specifically, the input of the well production model of the first well may be the parameter values of the plurality of operating parameters obtained by the first parameter value obtaining unit 100 and the second parameter value obtaining unit 200, and the output of the well production model of the first well may be the well production of the first well. The invention can also obtain the influence weight of each working parameter on the oil well yield of the first oil well after the machine learning is carried out on the training data. The influence weight of a certain working parameter is a quantitative index of the oil well yield change caused by the working parameter change, and when the influence weight of the working parameter is larger, the oil well yield change caused by the working parameter change is larger, and vice versa, the oil well yield change is smaller.

The device for determining the oil well yield provided by the embodiment of the invention can obtain the parameter values of various working parameters of the first oil well through the sensors and the indicator diagram respectively, and input the parameter values into the oil well yield model of the first oil well to obtain the oil well yield of the first oil well. Because the oil well production model is a model of a single oil well and is obtained by machine learning according to the training data of the oil well, the oil well production model used by the invention has stronger pertinence to the single oil well. Therefore, the accuracy of the oil well yield determined by the method is higher. Meanwhile, the oil well yield is obtained through the oil well yield model obtained through machine learning, so that the method does not need to perform complicated operation on the oil well yield, and is more convenient and quicker.

Specifically, when the oil well yield model is established, the embodiment of the invention can perform butt joint modeling on each working parameter in the training data and the oil well yield, so as to determine the matching relationship between the working parameter and the oil well yield. Specifically, the matching relationship may be embodied in a manner of affecting the weight. Namely: the oil well yield model can also obtain the following results after machine learning is carried out on training data: a weight of an effect of each of the operating parameters on well production from the first well. On this basis, the apparatus shown in fig. 4 may further include: a weight output unit for outputting the influence weight.

Further, the first yield obtaining unit 300 in the apparatus shown in fig. 4 may be specifically configured to: and inputting the obtained parameter values into a well production model of the first well, and obtaining the well production of the first well determined and output by the well production model according to the influence weight and the input parameter values, wherein the well production model of the first well is a model for outputting the well production of the first well, which is obtained by machine learning of the training data of the first well.

Optionally, another oil well production determining apparatus provided in the embodiment of the present invention may further include: a second yield obtaining unit and an error determining unit,

the second yield obtaining unit is used for obtaining the actual oil well yield of the first oil well in the first time period;

and the error determining unit is used for determining the error of the oil well yield output by the oil well yield model of the first oil well relative to the actual oil well yield, and when the error is larger than a threshold value, machine learning is carried out on each obtained parameter value and the actual oil well yield so as to adjust the oil well yield model of the first oil well.

The modification or addition of the training data does not result in a change to the machine learning algorithm of the well production model, nor does the present invention require modification of the machine learning algorithm of the well production model. Since the modification and addition of training data is very simple for the technician, the technician does not need to know the machine learning algorithm, greatly reducing the requirements on the technician. For example: the oil well production determining device provided by the embodiment of the invention is composed of a client device and a server which are connected in a communication mode, wherein a first parameter value obtaining unit 100 and a second parameter value obtaining unit 200 are located in the client device, and the client device sends obtained parameter values to the server. The first yield acquisition unit 300 is located in the server. Further, the server may also return the resulting well production to the client device, which may output it to inform the client. Therefore, in the process, the user only needs to input all the working parameters into the client device, and the method is simple and convenient. Meanwhile, in the training process of the oil well yield model, the client device sends the training data to the server, and the server performs machine learning on the training data. When a user wants to optimize the oil well yield model, the user only needs to input each new training data into the client device, and does not need to know the specific machine learning process, so that the operation burden of the user is greatly reduced.

According to the embodiment of the invention, when the error of the oil well yield output by the oil well yield model relative to the actual oil well yield is larger than the threshold value, the parameter values of the working parameters and the actual oil well yield in the first time period are used as training data, and the oil well yield model is controlled to perform machine learning on the training data. Through the machine learning, the oil well yield model can be optimized, and therefore the accuracy of the output oil well yield is improved.

The oil well production determining device comprises a processor and a memory, wherein the first parameter value obtaining unit, the second parameter value obtaining unit, the first production obtaining unit and the like are stored in the memory as program units, and the processor executes the program units stored in the memory to realize corresponding functions.

The processor comprises a kernel, and the kernel calls the corresponding program unit from the memory. The kernel can be set to one or more, and the oil well yield is determined by adjusting the kernel parameters.

The memory may include volatile memory in a computer readable medium, Random Access Memory (RAM) and/or nonvolatile memory such as Read Only Memory (ROM) or flash memory (flash RAM), and the memory includes at least one memory chip.

An embodiment of the present invention provides a storage medium having a program stored thereon, which when executed by a processor, implements the well production determination method.

The embodiment of the invention provides a processor for running a program, wherein the program is run to execute the oil well production determination method.

The embodiment of the invention provides equipment, which comprises a processor, a memory and a program which is stored on the memory and can run on the processor, wherein the processor executes the program and realizes the following steps:

a method of determining well production, comprising:

Optionally, the method further includes:

and outputting the influence weight.

The device herein may be a server, a PC, a PAD, a mobile phone, etc.

The present application further provides a computer program product adapted to perform a program for initializing the following method steps when executed on a data processing device:

a method of determining well production, comprising:

Optionally, the method further includes:

and outputting the influence weight.

As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory may include forms of volatile memory in a computer readable medium, Random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). The memory is an example of a computer-readable medium.

Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.

It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in the process, method, article, or apparatus that comprises the element.

The above are merely examples of the present application and are not intended to limit the present application. Various modifications and changes may occur to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A method of determining production from an oil well, comprising:

and selecting a well yield model matched with the acquired parameter values from the well yield model set of the first well according to the acquired parameter values, inputting the acquired parameter values into the well yield model of the first well, and acquiring the well yield of the first well output by the well yield model, wherein the well yield model of the first well is a model for outputting the well yield of the first well, which is acquired by machine learning of training data of the first well, and after a plurality of well yield models of the same well are acquired, model matching conditions respectively corresponding to the well yield models of the same well are stored, and the model matching conditions are used for determining the well yield model matched with the acquired parameter values of the working parameters of the well.

2. The method of claim 1, wherein the training data for the first well comprises: the production of the first oil well in a historical time period, the parameter values of the various operating parameters of the first oil well in the historical time period obtained by various sensors arranged on the oil production equipment of the first oil well, and the parameter values of the various operating parameters of the first oil well in the historical time period obtained according to an indicator diagram of the oil production equipment of the first oil well;

3. The method of claim 1 or 2, wherein the oil production equipment is a pumping unit, and the plurality of operating parameters obtained by the plurality of sensors comprises: at least one of indicator diagram load, pumping unit stroke, oil pressure, casing pressure, working current of the motor of the pumping unit and working voltage of the motor of the pumping unit,

4. The method of claim 1, further comprising:

obtaining actual well production for the first well over the first time period;

5. The method of claim 1, wherein the machine learning of the training data further results in: a weight of an effect of each of the operating parameters on well production from the first well, the method further comprising:

and outputting the influence weight.

6. The method of claim 5, wherein the selecting, from the set of well production models of the first well, a well production model matched with the obtained parameter values, and inputting the obtained parameter values into the well production model of the first well to obtain the well production of the first well output by the well production model, comprises:

in the oil well yield model set of the first oil well, selecting an oil well yield model matched with the first oil well according to each obtained parameter value;

7. An oil well production determination apparatus, comprising: a first parameter value obtaining unit, a second parameter value obtaining unit, and a first yield obtaining unit,

the first yield obtaining unit is configured to select, in the well yield model set of the first well, a well yield model matched with each obtained parameter value according to the obtained parameter value, input each obtained parameter value into the well yield model of the first well, and obtain the well yield of the first well output by the well yield model, where the well yield model of the first well is a model for outputting the well yield of the first well obtained after machine learning of the training data of the first well, and after obtaining multiple well yield models of the same well, store model matching conditions corresponding to the well yield models of the same well, and the model matching conditions are used to determine the well yield model matched with the obtained parameter value of the operating parameter of the well.

8. The apparatus of claim 7, wherein the oil production device is a pumping unit, and the plurality of operating parameters obtained by the plurality of sensors comprises: at least one of indicator diagram load, pumping unit stroke, oil pressure, casing pressure, working current of the motor of the pumping unit and working voltage of the motor of the pumping unit,

9. A storage medium comprising a stored program, wherein the program, when executed, controls an apparatus in which the storage medium is located to perform the method of determining well production according to any one of claims 1 to 6.

10. A processor for running a program, wherein the program when run performs the method of determining well production as defined in any one of claims 1 to 6.