An intelligent petroleum proppant raw material conveying equipment

Abstract

The present invention discloses an intelligent petroleum proppant raw material conveying equipment, which relates to the technical field of petroleum exploitation, and comprises a conveying mechanism for conveying the petroleum proppant raw material to the interior of an oil field. A control module is loaded on the conveying mechanism, and the control module collects a first historical training data set in advance. Based on the first historical training data set, the control module trains a first machine learning model for predicting the injection flow value of the petroleum proppant raw material in the conveying mechanism; the control module collects injection pressure data, formation pressure data and formation temperature data of the conveying mechanism in real time; the flow rate of the petroleum proppant raw material injected into the oil wellbore by the conveying mechanism is precisely controlled, and it can be ensured that the petroleum proppant raw material is injected into the oil reservoir at an optimal rate, thereby achieving the best formation coverage and support effect, which is helpful to improve the production efficiency of the oil reservoir and increase the output.

Description

Technical Field

[0001] The invention relates to the technical field of oil extraction, and in particular to intelligent oil proppant raw material conveying equipment. Background Art

[0002] Petroleum proppant is a special chemical used to improve oil well productivity and increase oil recovery. It is usually injected into the oil reservoir to increase the pressure of the formation and improve the fluidity of crude oil. It mainly includes base liquid and polymer particles, which can form a network structure in the oil reservoir or increase the viscosity of the fluid, thereby improving the fluidity of crude oil in the formation.

[0003] Existing intelligent petroleum proppant raw material conveying equipment, such as the carbon dioxide fracturing sand adding equipment and petroleum machinery disclosed in the Chinese patent with authorization announcement number CN217354341U, cannot make timely adjustments according to changes in pressure, temperature, etc. in the oil reservoir during the process of injecting petroleum proppant into the oil reservoir, resulting in the flow rate and concentration of the petroleum proppant cannot be accurately controlled. Summary of the invention

[0004] In order to overcome the above-mentioned technical problems, the purpose of the present invention is to provide an intelligent petroleum proppant raw material conveying equipment to solve the problem in the prior art that, during the process of injecting petroleum proppant into an oil reservoir, the existing conveying equipment cannot accurately control the flow rate and concentration of the petroleum proppant according to changes in the pressure, temperature, etc. in the oil reservoir, resulting in the petroleum proppant supporting effect in the oil reservoir failing to meet the expected problem.

[0005] The purpose of the present invention can be achieved through the following technical solutions:

[0006] Specifically, an intelligent petroleum proppant raw material conveying device is provided, including a conveying mechanism, which is used to convey the petroleum proppant raw material to the inside of the oil field. A control module is loaded on the conveying mechanism, and the control module collects a first historical training data set in advance. Based on the first historical training data set, the control module trains a first machine learning model for predicting the injection flow value of the petroleum proppant raw material in the conveying mechanism; the control module collects injection pressure data, formation pressure data and formation temperature data of the conveying mechanism in real time; based on the injection pressure data, formation pressure data and formation temperature data, the control module uses the first machine learning model to predict the injection flow value of the petroleum proppant raw material in the conveying mechanism; the control module controls the conveying mechanism based on the injection flow value;

[0007] The control module collects a second historical training data set in advance, and based on the second historical training data set, the control module trains a second machine learning model for predicting the injection liquid concentration value of the petroleum proppant raw material in the conveying mechanism;

[0008] The control module collects oil-water ratio data, formation temperature data and injection flow rate value predicted by the first machine learning model of the conveying mechanism in real time, and based on the oil-water ratio data, formation temperature data and injection flow rate value, the control module uses the second machine learning model to predict the injection liquid concentration value of the petroleum proppant raw material in the conveying mechanism;

[0009] The control module controls the delivery mechanism based on the injected liquid concentration value.

[0010] As a further solution of the present invention: the historical first training data set includes N groups of first training data, N is a positive integer, the first training data includes first feature data and first label data, wherein the first feature data includes injection pressure data, formation pressure data and formation temperature data;

[0011] The injection pressure data is the pressure value of the petroleum proppant raw material in the conveying mechanism entering the oil wellbore when collecting each set of first training data;

[0012] The formation pressure data is the formation pressure value in the reservoir when each set of first training data is collected;

[0013] The formation temperature data is the formation temperature value in the reservoir when each set of first training data is collected;

[0014] The first label data is the injection flow value of the petroleum proppant raw material in the conveying mechanism when collecting each set of first training data.

[0015] As a further solution of the present invention, the control module trains a first machine learning model for predicting the injection flow rate value of the petroleum proppant raw material in the conveying mechanism in the following manner:

[0016] The first feature data in each group of first training data is converted into a first feature vector, each group of first feature vectors is used as the input of the first machine learning model, the injection flow value predicted by the first machine learning model for each group of first feature data is used as the output, the injection flow value in the first label data corresponding to each group of first feature data is used as the prediction target, and minimizing the sum of the first prediction accuracies of all first feature vectors is used as the training target;

[0017] Among them, the calculation formula for the first prediction accuracy is ai=(bi-ci) 2 , where ai is the first prediction accuracy, bi is the predicted injection flow value corresponding to the first feature vector of the i-th group, and ci is the injection flow value in the i-th group of first label data. The first machine learning model is trained until the sum of the first prediction accuracies reaches convergence and the training is stopped.

[0018] As a further solution of the present invention: the conveying mechanism includes an input pipeline and a stirring box arranged on one side of the input pipeline;

[0019] The top end of the input pipe is bolted to a driving motor 1, the output shaft of the driving motor 1 passes through the input pipe and is connected to a screw rod, and the bottom end of the input pipe extends into the oil wellbore;

[0020] A second driving motor is bolted to the middle position of the top surface of the mixing box, a liquid delivery pipe is connected to one side of the top surface of the mixing box, and a particle delivery pipe is connected to the other side of the top surface of the mixing box. The output shaft of the second driving motor passes through the mixing box and is connected to the mixing shaft. One side of the bottom surface of the mixing box is connected to the top of the side of the input pipe through a pipe.

[0021] As a further solution of the present invention: a first pressure sensor is installed at the bottom end of the inner cavity of the input pipeline, and the first pressure sensor is used to collect the pressure value of the petroleum proppant raw material entering the oil wellbore.

[0022] As a further solution of the present invention: the control module is connected to a second pressure sensor and a temperature sensor via a wireless network, and the second pressure sensor and the temperature sensor are arranged in a stratum at the bottom of the oil wellbore;

[0023] The second pressure sensor is used to collect the formation pressure value in the oil reservoir, and the temperature sensor is used to collect the formation temperature value in the oil reservoir.

[0024] As a further solution of the present invention: the historical second training data set includes M groups of second training data, M is a positive integer, each group of second training data includes second feature data and second label data, wherein the second feature data includes oil-water ratio data, formation permeability data, formation temperature data and injection flow value;

[0025] The oil-water ratio data is the volume ratio of crude oil and water produced by the oil well when collecting each set of second training data;

[0026] The formation permeability data is the permeability of the petroleum proppant raw material in the formation when collecting each set of second training data;

[0027] The formation temperature data is the formation temperature value in the reservoir when each set of second training data is collected;

[0028] The injection flow value is the injection flow value predicted by the control module using the first machine learning model when collecting each set of second training data;

[0029] The second label data is the injected liquid concentration value of the petroleum proppant raw material in the conveying mechanism when each set of second training data is collected.

[0030] As a further solution of the present invention, the control module trains a second machine learning model for predicting the injection liquid concentration value of the petroleum proppant raw material in the conveying mechanism in the following manner:

[0031] The second feature data in each group of second training data is converted into a second feature vector, each group of second feature vectors is used as the input of the second machine learning model, the injection liquid concentration value predicted by the second machine learning model for each group of second feature data is used as the output, the injection liquid concentration value in the second label data corresponding to each group of second feature data is used as the prediction target, and minimizing the sum of the second prediction accuracies of all second feature vectors is used as the training target;

[0032] Among them, the calculation formula for the second prediction accuracy is xj=(yj-zj) 2 , where xj is the second prediction accuracy, yj is the predicted injection liquid concentration value corresponding to the j-th group of second eigenvectors, and zj is the injection liquid concentration value in the j-th group of second label data. The second machine learning model is trained until the sum of the second prediction accuracies converges and the training is stopped.

[0033] As a further solution of the present invention: the permeability of the petroleum proppant raw material in the formation is calculated as follows:

[0034] k=(Q×L×μ) / (A×ΔP);

[0035] Wherein, k is the permeability, Q is the injection flow value of the petroleum proppant raw material, L is the length of the formation, μ is the viscosity of the petroleum proppant raw material, A is the cross-sectional area of ​​the formation, and ΔP is the pressure drop of the petroleum proppant raw material in the formation.

[0036] Beneficial effects of the present invention:

[0037] 1. In the present invention, through the control module, when the conveying mechanism conveys the petroleum proppant raw material to the inside of the oil field, the control module collects the injection pressure data, formation pressure data and formation temperature data of the conveying mechanism in real time, and then uses the first machine learning model to predict the injection flow value of the petroleum proppant raw material in the conveying mechanism based on the injection pressure data, formation pressure data and formation temperature data. Finally, the control module controls the conveying mechanism according to the injection flow value predicted by the first machine learning model, so that the flow rate of the petroleum proppant raw material injected into the oil wellbore by the conveying mechanism is accurately controlled, which can ensure that the petroleum proppant raw material is injected into the oil reservoir at an optimal rate, thereby achieving the best formation coverage and support effect, which is helpful to improve the production efficiency of the oil reservoir and increase the output;

[0038] The flow rate of petroleum proppant raw materials is precisely controlled to avoid excessive or insufficient injection, reduce the waste of petroleum proppant, and save costs and resources;

[0039] Precise control of the flow of proppant feedstock can maximize reservoir productivity and extend reservoir life. It can also reduce operational risks and problems caused by over-injection or under-injection, such as formation damage and wellbore blockage, and help ensure safe production in oil fields.

[0040] 2. In the present invention, through the control module that is set up, in the process of the conveying mechanism conveying the petroleum proppant raw material to the interior of the oil field, the control module collects oil-water ratio data, formation permeability data, formation temperature data and injection flow value, and then uses the second machine learning model to predict the injection liquid concentration value of the petroleum proppant raw material in the conveying mechanism, so that the quality of the polymer particles in the petroleum proppant raw material can be accurately controlled, ensuring that the petroleum proppant raw material is injected into the oil reservoir at an optimal concentration, thereby achieving the best formation coverage and support effect, which helps to improve the production efficiency of the oil reservoir and increase output.

[0041] 3. In the present invention, by setting up the dividing mechanism, when the dividing wheel rotates, the control tube inside the dividing wheel does not rotate, and when the connecting channel rotates to the position of the negative pressure chamber, the negative pressure chamber can cause the connecting channel to generate suction, which can promote the polymer particles in the material box to enter the dividing trough, and when the connecting channel rotates to the position of the high pressure chamber, the high pressure chamber can cause the connecting channel to generate thrust, which can promote the discharge efficiency and effect of the polymer particles in the dividing trough and prevent the polymer particles from remaining. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The present invention will be further described below in conjunction with the accompanying drawings.

[0043] Figure 1 is a flowchart of the control module in Embodiment 1 of the present invention;

[0044] Figure 2 is a flowchart of the control module in Embodiment 2 of the present invention;

[0045] Figure 3 It is a schematic diagram of the internal structure of the conveying mechanism in the present invention;

[0046] Figure 4 It is a schematic diagram of the internal structure of the material box in the present invention;

[0047] Figure 5 It is a schematic diagram of the internal structure of the material dividing wheel in the present invention.

[0048] In the figure: 1. input pipeline; 11. driving motor 1; 12. screw; 2. stirring box; 21. driving motor 2; 22. stirring shaft; 23. liquid conveying pipe; 24. particle conveying pipe; 3. material box; 31. material distribution wheel; 311. control pipe; 312. negative pressure chamber; 313. high pressure chamber; 314. air delivery hole; 315. connecting channel; 32. material distribution trough. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] Embodiment 1:

[0051] like Figure 1 and Figure 3 As shown, the present invention discloses an intelligent petroleum proppant raw material conveying equipment, including a conveying mechanism, which is used to convey the petroleum proppant raw material to the inside of the oil field. It should be noted that the conveying mechanism is arranged at the top of the oil wellbore, and then the petroleum proppant raw material is conveyed to the bottom of the oil wellbore by high pressure, and then conveyed to the formation by the oil wellbore to fill the gaps in the formation, ensure the permeability effect of the formation, and improve the oil yield of the formation. A control module is loaded on the conveying mechanism. It should be noted that a microcontroller can be arranged on the conveying mechanism, and the control module is loaded on the microcontroller. A wireless network module is also arranged on the microcontroller. The control module can be connected to a back-end platform through the wireless network module. The back-end platform can be a PC terminal for monitoring the conveying mechanism. The staff can directly monitor and adjust the control module through the PC terminal, and indirectly adjust the conveying mechanism through the control module. The control module collects the first historical training data set in advance. Based on the first historical training data set, the control module trains a first machine learning model for predicting the injection flow value of the petroleum proppant raw material in the conveying mechanism; the control module collects the injection pressure data, formation pressure data and formation temperature data of the conveying mechanism in real time;

[0052] The historical first training data set includes N groups of first training data, where N is a positive integer, and the first training data includes first feature data and first label data, wherein the first feature data includes injection pressure data, formation pressure data, and formation temperature data. It should be noted that the historical first training data set is collected by technicians in the process of injecting petroleum proppant raw materials into oil reservoirs;

[0053] The injection pressure data is the pressure value of the oil proppant raw material in the conveying mechanism entering the oil wellbore when collecting each set of the first training data. It should be noted that a pressure sensor is provided in the conveying mechanism, and the setting position of the pressure sensor is adaptively adjusted by a person skilled in the art according to the specifications and internal space of the conveying mechanism. The pressure sensor can collect the pressure value of the oil proppant raw material in the conveying mechanism, and then convert the pressure value into an electrical signal and transmit it to the control module, and the control module performs corresponding processing on the transmitted electrical signal;

[0054] The formation pressure data is the formation pressure value in the oil reservoir when collecting each set of the first training data. It should be noted that a pressure sensor can be installed at the bottom of the oil wellbore. The pressure sensor set when the oil wellbore is arranged can be used. The pressure sensor directly collects the formation pressure value in the oil reservoir at the bottom of the oil wellbore, and then converts the formation pressure value in the oil reservoir into an electrical signal and transmits it to the control module. The control module processes the transmitted electrical signal accordingly.

[0055] The formation temperature data is the formation temperature value in the oil reservoir when collecting each set of the first training data. It should be noted that a temperature sensor can be installed at the bottom of the oil wellbore. The temperature sensor set when the oil wellbore is arranged can be used. The temperature sensor directly collects the formation temperature value in the oil reservoir at the bottom of the oil wellbore, and then converts the formation temperature value in the oil reservoir into an electrical signal and transmits it to the control module. The control module performs corresponding processing on the transmitted electrical signal;

[0056] The first label data is the injection flow value of the petroleum proppant raw material in the conveying mechanism when collecting each set of the first training data. It should be noted that a flow meter can be set at the junction of the conveying mechanism and the oil wellbore. The flow meter directly collects the injection flow value of the petroleum proppant raw material entering the oil wellbore, and then converts the injection flow value of the petroleum proppant raw material into an electrical signal and transmits it to the control module. The control module performs corresponding processing on the transmitted electrical signal.

[0057] The control module trains a first machine learning model for predicting the injection flow value of the petroleum proppant raw material in the conveying mechanism in the following manner:

[0058] The first feature data in each group of first training data is converted into a first feature vector, each group of first feature vectors is used as the input of the first machine learning model, the injection flow value predicted by the first machine learning model for each group of first feature data is used as the output, the injection flow value in the first label data corresponding to each group of first feature data is used as the prediction target, and minimizing the sum of the first prediction accuracies of all first feature vectors is used as the training target;

[0059] Among them, the calculation formula for the first prediction accuracy is ai=(bi-ci) 2 , where ai is the first prediction accuracy, bi is the predicted injection flow value corresponding to the first feature vector of the i-th group, and ci is the injection flow value in the i-th group of first label data. The first machine learning model is trained until the sum of the first prediction accuracies reaches convergence and the training is stopped. The convergence standard is set by those skilled in the art according to the specific model training situation;

[0060] It should be noted that, through the control module, when the conveying mechanism conveys the petroleum proppant raw material to the inside of the oil field, the control module collects the injection pressure data, formation pressure data and formation temperature data of the conveying mechanism in real time. It should also be noted that a clock module is also provided on the conveying mechanism, and the control module controls the frequency of collecting the injection pressure data, formation pressure data and formation temperature data according to the clock module;

[0061] The technicians in this field make corresponding adjustments according to the specifications of the oil reservoir and the type of formation. For example, the control module collects data at a frequency of 1 second / time, that is, the control module collects injection pressure data, formation pressure data and formation temperature data once every 1 second, and then uses the first machine learning model to predict the injection flow value of the petroleum proppant raw material in the conveying mechanism based on the injection pressure data, formation pressure data and formation temperature data. Finally, the control module controls the conveying mechanism according to the injection flow value predicted by the first machine learning model, so that the flow rate of the petroleum proppant raw material injected into the oil wellbore by the conveying mechanism is accurately controlled, which can ensure that the petroleum proppant raw material is injected into the oil reservoir at an optimal rate, thereby achieving the best formation coverage and support effect, which is helpful to improve the production efficiency of the oil reservoir and increase the output;

[0062] The flow rate of petroleum proppant raw materials is precisely controlled to avoid excessive or insufficient injection, reduce the waste of petroleum proppant, and save costs and resources;

[0063] Precise control of the flow of proppant feedstock can maximize reservoir productivity and extend reservoir life. It can also reduce operational risks and problems caused by over-injection or under-injection, such as formation damage and wellbore blockage, and help ensure safe production in oil fields.

[0064] like Figure 2 and Figure 3 As shown, the conveying mechanism includes an input pipeline 1 and a mixing box 2 arranged on one side of the input pipeline 1; the top of the input pipeline 1 is bolted to a driving motor 11, the output shaft of the driving motor 11 passes through the input pipeline 1 and is connected to a screw 12, and the bottom end of the input pipeline 1 extends into the oil wellbore; the middle position of the top surface of the mixing box 2 is bolted to a driving motor 21, one side of the top surface of the mixing box 2 is connected to a liquid conveying pipe 23, the other side of the top surface of the mixing box 2 is connected to a particle conveying pipe 24, the output shaft of the driving motor 21 passes through the mixing box 2 and is connected to a mixing shaft 22, and one side of the bottom surface of the mixing box 2 is connected to the top of the side of the input pipeline 1 through a pipeline;

[0065] It should be noted that the liquid delivery pipe 23 is used to connect the pipeline of the base liquid of the petroleum proppant raw material. The base liquid can be water, oil or other solvents. The selection of the base liquid depends on the application requirements and geological conditions, and is adaptively adjusted by technical personnel in the field according to their own experience and needs. The technical personnel in the field can also add surfactants, cross-linking agents, sustained-release agents, pH regulators, etc. to the base liquid according to their own experience and needs. The particle delivery pipe 24 is used to connect the pipeline of polymer particles. The polymer particles can form a network structure in the oil well reservoir, improve the permeability of the reservoir, and prevent the decline of oil well production capacity.

[0066] A first pressure sensor is installed at the bottom of the inner cavity of the input pipe 1. The first pressure sensor is used to collect the pressure value of the petroleum proppant raw material entering the oil wellbore, and then convert the pressure value into an electrical signal and transmit it to the control module. The control module processes the transmitted electrical signal accordingly.

[0067] The control module is connected to a second pressure sensor and a temperature sensor through a wireless network. The second pressure sensor and the temperature sensor are arranged in the stratum at the bottom of the oil wellbore. Then, the stratum pressure value in the oil reservoir is converted into an electrical signal and transmitted to the control module. The control module processes the transmitted electrical signal accordingly.

[0068] The second pressure sensor is used to collect the formation pressure value in the oil reservoir, and the temperature sensor is used to collect the formation temperature value in the oil reservoir. The formation temperature value in the oil reservoir is then converted into an electrical signal and transmitted to the control module, and the control module performs corresponding processing on the transmitted electrical signal.

[0069] The control module collects a second historical training data set in advance, and based on the second historical training data set, the control module trains a second machine learning model for predicting the injection liquid concentration value of the petroleum proppant raw material in the conveying mechanism, wherein the second historical training data set is collected by technicians in the process of injecting the petroleum proppant raw material into the oil reservoir;

[0070] The control module collects oil-water ratio data, formation temperature data and injection flow rate value predicted by the first machine learning model of the conveying mechanism in real time, and based on the oil-water ratio data, formation temperature data and injection flow rate value, the control module uses the second machine learning model to predict the injection liquid concentration value of the petroleum proppant raw material in the conveying mechanism;

[0071] The control module controls the delivery mechanism based on the injected liquid concentration value.

[0072] The historical second training data set includes M groups of second training data, where M is a positive integer, and each group of second training data includes second feature data and second label data, wherein the second feature data includes oil-water ratio data, formation permeability data, formation temperature data, and injection flow value;

[0073] The oil-water ratio data is the volume ratio of crude oil and water produced by the oil well when collecting each set of second training data. It should be noted that the volume ratio value is collected by a person skilled in the art using a conductivity or resistivity sensor. Since crude oil and water have significant differences in conductivity or resistivity, the conductivity or resistivity sensor can be used to distinguish whether the flow is water or oil, and estimate their ratio. The person skilled in the art obtains the volume ratio of crude oil and water by measuring the conductivity or resistivity of the fluid;

[0074] The formation permeability data is the permeability of the petroleum proppant raw material in the formation when collecting each set of second training data. It should be noted that:

[0075] The permeability of petroleum proppant raw materials in the formation is calculated as follows:

[0076] k=(Q×L×μ) / (A×ΔP);

[0077] Wherein, k is permeability, Q is the injection flow value of the petroleum proppant raw material, L is the length of the formation, μ is the viscosity of the petroleum proppant raw material, A is the cross-sectional area of ​​the formation, and ΔP is the pressure drop of the petroleum proppant raw material in the formation. Pressure drop refers to the pressure loss caused by resistance during the flow of fluid, and is usually used to describe the change in pressure when the fluid passes through a pipeline, a pipeline element or a formation. When designing a pipeline system or predicting the production of an oil well, the technicians in this field can use the principles of fluid mechanics and the pressure loss model to calculate the expected pressure drop;

[0078] The formation temperature data is the formation temperature value in the oil reservoir when collecting each set of second training data. It should be noted that a temperature sensor can be installed at the bottom of the oil wellbore. The temperature sensor set when the oil wellbore is arranged can be used. The temperature sensor directly collects the formation temperature value in the oil reservoir at the bottom of the oil wellbore, and then converts the formation temperature value in the oil reservoir into an electrical signal and transmits it to the control module. The control module performs corresponding processing on the transmitted electrical signal;

[0079] The injection flow value is the injection flow value predicted by the control module using the first machine learning model when collecting each set of second training data. It should be noted that the injection flow value is the flow rate of the petroleum proppant raw material injected into the oil wellbore by the conveying mechanism;

[0080] The second label data is the injected liquid concentration value of the petroleum proppant raw material in the conveying mechanism when each set of second training data is collected.

[0081] The control module trains a second machine learning model to predict the injection liquid concentration value of the petroleum proppant raw material in the conveying mechanism as follows:

[0082] The second feature data in each group of second training data is converted into a second feature vector, each group of second feature vectors is used as the input of the second machine learning model, the injection liquid concentration value predicted by the second machine learning model for each group of second feature data is used as the output, the injection liquid concentration value in the second label data corresponding to each group of second feature data is used as the prediction target, and minimizing the sum of the second prediction accuracies of all second feature vectors is used as the training target;

[0083] Among them, the calculation formula for the second prediction accuracy is xj=(yj-zj) 2 , where xj is the second prediction accuracy, yj is the predicted injection liquid concentration value corresponding to the j-th group of second feature vectors, and zj is the injection liquid concentration value in the j-th group of second label data. The second machine learning model is trained until the sum of the second prediction accuracies reaches convergence and the training is stopped. The convergence standard is set by those skilled in the art according to the specific model training situation;

[0084] It should be noted that the injection liquid concentration value refers to the mass ratio of polymer particles to base liquid in the petroleum proppant raw material. The specific calculation formula is injection liquid concentration value = (m 溶质 / m 溶液 )×100%;

[0085] Where m 溶质 Refers to the mass of aggregate particles, m 溶液 is the mass of the base fluid;

[0086] The mass of aggregate particles and the mass of base liquid are collected by:

[0087] A flow meter is provided inside the liquid delivery pipe 23, and the flow meter collects the flow rate Q of the base liquid, and then:

[0088] m 溶液 = Q × t × ρ;

[0089] t is the unit time, i.e., the time for the liquid delivery pipe 23 to pass through the base liquid, and ρ is the density of the base liquid;

[0090] It should be noted that, through the set control module, in the process of the conveying mechanism conveying the petroleum proppant raw material to the interior of the oil field, the control module collects oil-water ratio data, formation permeability data, formation temperature data and injection flow value, and then uses the second machine learning model to predict the injection liquid concentration value of the petroleum proppant raw material in the conveying mechanism, so that the quality of the polymer particles in the petroleum proppant raw material can be accurately controlled, ensuring that the petroleum proppant raw material is injected into the oil reservoir at an optimal concentration, thereby achieving the best formation coverage and support effect, which helps to improve the production efficiency of the oil reservoir and increase output.

[0091] Embodiment 2:

[0092] like Figure 3 , Figure 4 and Figure 5 As shown, one end of the particle conveying pipe 24 away from the mixing box 2 is connected to a distributing mechanism for controlling the conveying quality of the polymer particles, the distributing mechanism includes a material box 3, the polymer particles are stored inside the material box 3, a distributing wheel 31 is arranged at the bottom of the inner cavity of the material box 3, a plurality of distributing grooves 32 are opened on the side of the distributing wheel 31, one side of the distributing wheel 31 is connected to the particle conveying pipe 24, when the distributing wheel 31 rotates, the polymer particles stored inside the material box 3 will fall into the distributing groove 32 on the top of the distributing wheel 31, with the rotation of the distributing wheel 31, the distributing wheel 31 can convey the polymer particles to the inside of the particle conveying pipe 24, completing the transportation of the polymer particles, those skilled in the art can accurately control the quality of the polymer particles by controlling the number of times the distributing groove 32 conveys the polymer particles to the particle conveying pipe 24 per unit time;

[0093] It should also be noted that a control tube 311 is provided inside the material distribution wheel 31, and a negative pressure chamber 312 and a high pressure chamber 313 are provided inside the control tube 311. A plurality of gas delivery holes 314 are provided on the side of the control tube 311, and the plurality of gas delivery holes 314 are evenly distributed on the outside of the negative pressure chamber 312 and the high pressure chamber 313. A connecting channel 315 is provided inside the material distribution wheel 31, and the position of the connecting channel 315 corresponds to the position of the material distribution trough 32. A mesh groove is provided on the inner end surface of the material distribution trough 32, and the mesh groove is connected to the connecting channel 315. The diameter of the mesh groove is smaller than the diameter of the polymerized particles. The end of the connecting channel 315 close to the control tube 311 is enlarged, and the enlarged diameter is larger than the distance between the two gas delivery holes 314.

[0094] A vacuum pump is arranged outside the material distribution wheel 31. The position and specification of the vacuum pump are adaptively selected by a person skilled in the art according to the conditions on site. The output end of the vacuum pump is connected to the negative pressure chamber 312 through a pipeline, and the input end of the vacuum pump is connected to the high pressure chamber 313 through a pipeline. When the vacuum pump is turned on, the vacuum pump can suck the air in the negative pressure chamber 312 and discharge it into the high pressure chamber 313.

[0095] A driving motor is connected to the center of one side of the dividing wheel 31, and the driving motor can directly drive the dividing wheel 31 to rotate. It should also be noted that when the dividing wheel 31 rotates, the control tube 311 inside the dividing wheel 31 does not rotate, and when the connecting channel 315 rotates to the position of the negative pressure chamber 312, the negative pressure chamber 312 can cause the connecting channel 315 to generate suction, which can promote the polymer particles in the material box 3 to enter the dividing trough 32, and when the connecting channel 315 rotates to the position of the high pressure chamber 313, the high pressure chamber 313 can cause the connecting channel 315 to generate thrust, which can promote the discharge efficiency and effect of the polymer particles in the dividing trough 32 and prevent the polymer particles from remaining.

[0096] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. An intelligent petroleum proppant raw material conveying equipment, characterized in that: include: A conveying mechanism, used to convey the petroleum proppant raw material to the interior of the oil field; A control module, which is loaded on the conveying mechanism, the control module collects a first historical training data set in advance, and based on the first historical training data set, the control module trains a first machine learning model for predicting an injection flow value of a petroleum proppant raw material in the conveying mechanism; The control module collects injection pressure data, formation pressure data and formation temperature data of the conveying mechanism in real time; Based on the injection pressure data, the formation pressure data, and the formation temperature data, the control module uses a first machine learning model to predict an injection flow value of the petroleum proppant raw material in the conveying mechanism; The control module controls the delivery mechanism based on the injection flow value; The control module collects a second historical training data set in advance, and based on the second historical training data set, the control module trains a second machine learning model for predicting the injection liquid concentration value of the petroleum proppant raw material in the conveying mechanism; The control module collects oil-water ratio data, formation temperature data and injection flow rate value predicted by the first machine learning model of the conveying mechanism in real time, and based on the oil-water ratio data, formation temperature data and injection flow rate value, the control module uses the second machine learning model to predict the injection liquid concentration value of the petroleum proppant raw material in the conveying mechanism; The control module controls the delivery mechanism based on the concentration value of the injected liquid; The historical first training data set includes N groups of first training data, where N is a positive integer, and the first training data includes first feature data and first label data, wherein the first feature data includes injection pressure data, formation pressure data, and formation temperature data; The injection pressure data is the pressure value of the petroleum proppant raw material in the conveying mechanism entering the oil wellbore when collecting each set of first training data; The formation pressure data is the formation pressure value in the reservoir when each set of first training data is collected; The formation temperature data is the formation temperature value in the reservoir when each set of first training data is collected; The first label data is the injection flow rate value of the petroleum proppant raw material in the conveying mechanism when collecting each set of first training data; The control module trains a first machine learning model for predicting the injection flow rate value of the petroleum proppant raw material in the delivery mechanism in the following manner: The first feature data in each group of first training data is converted into a first feature vector, each group of first feature vectors is used as the input of the first machine learning model, the injection flow value predicted by the first machine learning model for each group of first feature data is used as the output, the injection flow value in the first label data corresponding to each group of first feature data is used as the prediction target, and minimizing the sum of the first prediction accuracies of all first feature vectors is used as the training target; Among them, the calculation formula for the first prediction accuracy is ai=(bi-ci) 2 , where ai is the first prediction accuracy, bi is the predicted injection flow value corresponding to the first feature vector of the i-th group, and ci is the injection flow value in the i-th group of first label data. The first machine learning model is trained until the sum of the first prediction accuracies reaches convergence and the training is stopped; The historical second training data set includes M groups of second training data, where M is a positive integer, and each group of second training data includes second feature data and second label data, wherein the second feature data includes oil-water ratio data, formation permeability data, formation temperature data, and injection flow value; The oil-water ratio data is the volume ratio of crude oil and water produced by the oil well when collecting each set of second training data; The formation permeability data is the permeability of the petroleum proppant raw material in the formation when collecting each set of second training data; The formation temperature data is the formation temperature value in the reservoir when each set of second training data is collected; The injection flow value is the injection flow value predicted by the control module using the first machine learning model when collecting each set of second training data; The second label data is the injected liquid concentration value of the petroleum proppant raw material in the conveying mechanism when collecting each set of second training data; The control module trains a second machine learning model for predicting the injection liquid concentration value of the petroleum proppant raw material in the conveying mechanism in the following manner: The second feature data in each group of second training data is converted into a second feature vector, each group of second feature vectors is used as the input of the second machine learning model, the injection liquid concentration value predicted by the second machine learning model for each group of second feature data is used as the output, the injection liquid concentration value in the second label data corresponding to each group of second feature data is used as the prediction target, and minimizing the sum of the second prediction accuracies of all second feature vectors is used as the training target; Among them, the calculation formula for the second prediction accuracy is xj=(yj-zj) 2 , where xj is the second prediction accuracy, yj is the predicted injection liquid concentration value corresponding to the j-th group of second eigenvectors, and zj is the injection liquid concentration value in the j-th group of second label data. The second machine learning model is trained until the sum of the second prediction accuracies converges and the training is stopped.

2. The intelligent petroleum proppant raw material conveying equipment according to claim 1 is characterized in that: The conveying mechanism comprises an input pipeline (1) and a stirring box (2) arranged on one side of the input pipeline (1); The top end of the input pipe (1) is bolted to a drive motor (11); the output shaft of the drive motor (11) passes through the input pipe (1) and is connected to a screw rod (12); the bottom end of the input pipe (1) extends into the oil wellbore; A second drive motor (21) is bolted to the middle of the top surface of the stirring box (2); a liquid delivery pipe (23) is connected to one side of the top surface of the stirring box (2); a particle delivery pipe (24) is connected to the other side of the top surface of the stirring box (2); an output shaft of the second drive motor (21) passes through the stirring box (2) and is connected to a stirring shaft (22); and one side of the bottom surface of the stirring box (2) is connected to the top of the side surface of the input pipe (1) via a pipe.

3. The intelligent petroleum proppant raw material conveying equipment according to claim 2 is characterized in that: A first pressure sensor is installed at the bottom end of the inner cavity of the input pipeline (1), and the first pressure sensor is used to collect the pressure value of the petroleum proppant raw material entering the oil wellbore.

4. The intelligent petroleum proppant raw material conveying equipment according to claim 1 is characterized in that: The control module is connected to a second pressure sensor and a temperature sensor via a wireless network, and the second pressure sensor and the temperature sensor are arranged in a stratum at the bottom of the oil wellbore; The second pressure sensor is used to collect the formation pressure value in the oil reservoir, and the temperature sensor is used to collect the formation temperature value in the oil reservoir.

5. The intelligent petroleum proppant raw material conveying equipment according to claim 1 is characterized in that: The permeability of the petroleum proppant raw material in the formation is calculated as follows: k=(Q×L×μ) / (A×ΔP); Wherein, k is the permeability, Q is the injection flow value of the petroleum proppant raw material, L is the length of the formation, μ is the viscosity of the petroleum proppant raw material, A is the cross-sectional area of ​​the formation, and ΔP is the pressure drop of the petroleum proppant raw material in the formation.

Images