Petroleum proppant particle forming control method and system based on machine learning

Abstract

The invention relates to the technical field of petroleum proppant molding. The invention discloses a petroleum proppant particle molding control method and system based on machine learning, comprising first obtaining a base material weight and a base material purpose, determining an initial additive type according to the base material purpose and a first preset rule, obtaining an additive weight according to the base material weight and the initial additive type, determining a waste material type according to the initial additive type and a second preset rule, and then obtaining a waste weight through the additive weight and the waste type; if the initial additive type is chromite or pyrolusite, determining the chromite or pyrolusite as a target additive type; if the initial additive type is dolomite, determining the target additive type based on a first processing cost and a second processing cost; selecting additives according to demand or purpose, thereby improving the processing quality of the proppant.

Description

Technical Field

[0001] The present invention relates to the technical field of petroleum proppant molding, and more specifically, to a petroleum proppant particle molding control method and system based on machine learning. Background Art

[0002] At present, hydraulic fracturing production-increasing technology is widely used in the oil and gas industry and has become one of the important means to increase the production of oil and gas wells. This technology mixes fracturing proppants into fracturing fluid and injects them into deep rock fractures to support the rock fractures to form high-permeability crude oil flow channels, thereby maintaining smooth passage of crude oil and maximizing the oil production potential of oil wells. Currently, commonly used fracturing proppants mainly include quartz sand and expanded clay fracturing proppants. Expanded clay fracturing proppants are based on bauxite and have the advantages of high strength and low price. They are widely used in deep oil and gas wells.

[0003] However, there is a problem of high sintering temperature in the preparation of fracturing proppants from bauxite, which leads to high processing costs. In order to solve this problem, additives need to be introduced to reduce the sintering temperature of bauxite fracturing proppants. However, current technologies often only focus on a specific additive and lack overall consideration, which limits the improvement of fracturing proppant processing quality. With the development of artificial intelligence technology, the existing technology has used artificial intelligence technology to select the optimal proppant according to the actual situation of oil and gas wells and the purpose of the proppant, but artificial intelligence technology is rarely applied to the production process of proppants.

[0004] For example, a Chinese patent with publication number CN115758851A provides a method for selecting multi-scale proppants for fractures in formations containing natural fractures. The patent classifies the fractures by obtaining the fracture extension conditions, and then selects the optimal proppant based on the classification conditions. However, the above patent only applies artificial intelligence technology to the actual use of proppants, and does not apply it to the production process of proppants. For example, in the production process of proppants, due to different uses of proppants or different demands for proppants, proppants are required to have a low breakage rate or a high pass rate. In the process of proppant molding, it is impossible to select additives according to demand or use, thereby improving the processing quality of the proppant.

[0005] In view of this, the present invention proposes a petroleum proppant particle forming control method and system based on machine learning to solve the above problems. Summary of the invention

[0006] In order to overcome the above-mentioned defects of the prior art, the present invention provides a petroleum proppant particle forming control method and system based on machine learning.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] The method for controlling the formation of petroleum proppant particles based on machine learning includes:

[0009] S10: Obtaining the weight of the base material and the purpose of the base material, and determining the type of initial additive according to the purpose of the base material and the first preset rule, the initial additive type including chromite, dolomite and pyrolusite;

[0010] S20: Inputting the weight of the base material and the initial additive type into a pre-built first machine learning model to obtain the weight of the additive;

[0011] S30: Determine the type of waste material according to the initial additive type and the second preset rule, input the weight of the additive and the type of waste material into a pre-built second machine learning model, and obtain the weight of the waste material;

[0012] S40: If the initial additive type is chromite or pyrolusite, determining chromite or pyrolusite as the target additive type;

[0013] S50: If the initial additive type is dolomite, the weight of the burnt-out material is calculated according to the weight of the additive corresponding to the dolomite and the weight of the waste, the first processing cost is calculated according to the weight of the additive corresponding to the dolomite and the weight of the burnt-out material, the second processing cost is calculated according to the weight of the additive corresponding to the pyrolusite, and the target additive type is determined based on the first processing cost and the second processing cost;

[0014] Methods for calculating the weight of burnt material based on the weight of additives corresponding to dolomite and the weight of waste include:

[0015] ;

[0016] Where Wfl is the weight of the burned objects, is the weight of additive corresponding to dolomite, is the waste weight corresponding to dolomite, are all weight coefficients.

[0017] The method for calculating the first processing cost according to the weight of additives corresponding to dolomite and the weight of the burned objects includes:

[0018] ;

[0019] Where IPC is the first processing cost, is the price per unit weight of dolomite, The price corresponding to the unit weight of burned items.

[0020] Further, the base material uses include high pressure resistance uses, permeability uses and low sintering uses, and the method for determining the initial additive type according to the base material uses and the first preset rule includes:

[0021] When the base material is used for high pressure resistance, the initial additive type is chromite;

[0022] When the base material is used for penetration, the initial additive type is pyrolusite;

[0023] When the base material application is low sintering application, the initial additive type is dolomite.

[0024] Furthermore, the method for constructing the first machine learning model includes:

[0025] A sample data set is obtained, wherein the sample data set includes a historical base material weight, a historical initial additive type, and a historical additive weight. The sample data set is divided into a sample training set and a sample test set, and a regression network is constructed. The historical base material weight and the historical initial additive type in the sample training set are used as input data of the regression network, and the historical additive weight in the sample training set is used as output data of the regression network. The regression network is trained to obtain an initial regression network for predicting real-time additive weight. The initial regression network is tested using the sample test set, and the regression network that satisfies a preset error value is output as a first machine learning model.

[0026] Further, the waste types include calcium oxide and ferric oxide, and the method for determining the waste type according to the initial additive type and the second preset rule includes:

[0027] When the initial additive type is chromite or pyrolusite, the waste type is ferric oxide;

[0028] When the initial additive type is dolomite, the waste type is calcium oxide.

[0029] Furthermore, the method for constructing the second machine learning model includes:

[0030] A sample data set is obtained, wherein the sample data set includes historical additive weights, historical waste types, and historical waste weights; the sample data set is divided into a sample training set and a sample test set; a regression network is constructed; the historical additive weights and historical waste types in the sample training set are used as input data of the regression network; the historical waste weights in the sample training set are used as output data of the regression network; the regression network is trained to obtain an initial regression network for predicting real-time waste weight; the initial regression network is tested using the sample test set; and a regression network that satisfies an error value less than a preset value is output as a second machine learning model.

[0031] Further, the method for calculating the second processing cost according to the weight of the additive corresponding to the pyrolusite includes:

[0032] ;

[0033] Where SPC is the second processing cost, is the weight of additive corresponding to pyrolusite, It is the price corresponding to the unit weight of pyrolusite.

[0034] Further, the method for determining the target additive type based on the first processing cost and the second processing cost includes:

[0035] When the first processing cost is less than the second processing cost, the dolomite corresponding to the first processing cost is used as the target additive type;

[0036] When the first processing cost is greater than or equal to the second processing cost, the pyrolusite corresponding to the second processing cost is used as the target additive type.

[0037] A petroleum proppant particle forming control system based on machine learning, which is used to implement the above-mentioned petroleum proppant particle forming control method based on machine learning, comprises:

[0038] Data acquisition module: used to obtain the weight of the base material and the purpose of the base material, and determine the initial additive type according to the purpose of the base material and the first preset rule, the initial additive type including chromite, dolomite and pyrolusite;

[0039] An additive weight determination module: used to input the base material weight and the initial additive type into a pre-built first machine learning model to obtain the additive weight;

[0040] A waste weight determination module: used to determine the type of waste according to the initial additive type and a second preset rule, and input the additive weight and the type of waste into a pre-built second machine learning model to obtain the weight of the waste;

[0041] The first type determination module is used to determine chromite or pyrolusite as the target additive type if the initial additive type is chromite or pyrolusite;

[0042] The second type determination module is used to calculate the weight of the burnt material according to the weight of the additive corresponding to the dolomite and the weight of the waste if the initial additive type is dolomite, calculate the first processing cost according to the weight of the additive corresponding to the dolomite and the weight of the burnt material, calculate the second processing cost according to the weight of the additive corresponding to the pyrolusite, and determine the target additive type based on the first processing cost and the second processing cost.

[0043] A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed, implements the above-mentioned machine learning-based petroleum proppant particle forming control method.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] The present invention first obtains the weight of the base material and the purpose of the base material, determines the initial additive type according to the purpose of the base material and a first preset rule, obtains the additive weight according to the weight of the base material and the initial additive type, determines the waste type according to the initial additive type and a second preset rule, and then obtains the waste weight through the additive weight and the waste type. If the initial additive type is chromite or pyrolusite, the chromite or pyrolusite is determined as the target additive type. If the initial additive type is dolomite, the target additive type is determined based on the first processing cost and the second processing cost. The additive is selected according to demand or purpose, thereby improving the processing quality of the proppant. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a flow chart of the petroleum proppant particle forming control method based on machine learning in the present invention;

[0047] Figure 2 It is a schematic diagram of the relationship between the crushing rate and the sintering temperature in the present invention;

[0048] Figure 3 A schematic diagram of a computer-readable storage medium in the present invention. 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] Example 1

[0051] See also Figure 1 As shown, this embodiment discloses a method for controlling the formation of petroleum proppant particles based on machine learning, comprising:

[0052] S10: Obtaining the weight of the base material and the purpose of the base material, and determining the type of initial additive according to the purpose of the base material and the first preset rule, the initial additive type including chromite, dolomite and pyrolusite;

[0053] It should be noted that the base material in this embodiment refers to the main material for preparing petroleum proppants, and the base material is preferably bauxite. The main components of bauxite are aluminum oxide and silicon dioxide. The weight of the base material can be directly obtained by the sensor in the mixing and molding equipment. The mixing and molding equipment is preferably a horizontal ribbon mixer. It can also be obtained by other means. This embodiment does not impose too many restrictions on this.

[0054] The base material uses include high pressure resistance uses, penetration uses, and low sintering uses. The method for determining the initial additive type according to the base material uses and the first preset rule includes:

[0055] When the base material is used for high pressure resistance, the initial additive type is chromite;

[0056] When the base material is used for penetration, the initial additive type is pyrolusite;

[0057] When the base material application is low sintering application, the initial additive type is dolomite.

[0058] It can be understood that high pressure resistance means that the oil proppant is used in a high pressure environment. Generally, the deeper the formation depth, the greater the pressure generated by the formation fracture, which means that the oil proppant has the advantage of low breakage rate under high pressure environment. Figure 2 To illustrate with an example, Figure 2 B is bauxite, P is pyrolusite, D is dolomite, and C is chromite. Figure 2 It is clear that B plus C refers to the petroleum proppant made of bauxite mixed with chromite, which has the lowest crushing rate. The penetration purpose refers to the formation of a good pore structure inside the finished petroleum proppant, so that oil and gas can flow quickly. B plus P refers to the petroleum proppant made of bauxite mixed with pyrolusite, which has mostly uniformly distributed nearly spherical closed pores inside, so that oil and gas can flow quickly. B plus D refers to the petroleum proppant made of bauxite mixed with dolomite, such as Figure 2 As shown, normally, the sintering temperature of bauxite without additives is about 1480 degrees, but the sintering temperature of bauxite with dolomite added is reduced to about 1350 degrees, so the low sintering application refers to the ability to be processed with lower processing difficulty.

[0059] S20: Inputting the weight of the base material and the initial additive type into a pre-built first machine learning model to obtain the weight of the additive;

[0060] The first machine learning model construction method includes:

[0061] A sample data set is obtained, wherein the sample data set includes a historical base material weight, a historical initial additive type, and a historical additive weight. The sample data set is divided into a sample training set and a sample test set, and a regression network is constructed. The historical base material weight and the historical initial additive type in the sample training set are used as input data of the regression network, and the historical additive weight in the sample training set is used as output data of the regression network. The regression network is trained to obtain an initial regression network for predicting real-time additive weight. The initial regression network is tested using the sample test set, and a regression network that satisfies a preset error value is output as a first machine learning model. The regression network is preferably a neural network model.

[0062] It is understandable that the weight of the additive is obtained through experiments conducted by personnel in this field. When the weight of the base material is the same, different initial additive types will lead to different additive weights. Therefore, the initial additive type and the additive weight correspond to each other.

[0063] S30: Determine the type of waste material according to the initial additive type and the second preset rule, input the weight of the additive and the type of waste material into a pre-built second machine learning model, and obtain the weight of the waste material;

[0064] Specifically, the waste types include calcium oxide and ferric oxide. The waste used in this embodiment is steel slag. The steel slag is screened to extract the main components of calcium oxide and ferric oxide in the steel slag. The main source of calcium oxide with a high content in the steel slag is quicklime added when adjusting the alkalinity during steelmaking. It can be understood that since chromite and pyrolusite both have ferric oxide as one of their main components, and dolomite has calcium oxide as one of its main components, by adding ferric oxide or calcium oxide, chromite, pyrolusite or dolomite can be replaced to a certain extent, which not only recycles the waste, but also reduces the cost of petroleum proppant processing.

[0065] The method for determining the type of waste material according to the initial additive type and the second preset rule includes:

[0066] When the initial additive type is chromite or pyrolusite, the waste type is ferric oxide;

[0067] When the initial additive type is dolomite, the waste type is calcium oxide.

[0068] The second machine learning model construction method includes:

[0069] A sample data set is obtained, wherein the sample data set includes historical additive weight, historical waste type and historical waste weight, the sample data set is divided into a sample training set and a sample test set, a regression network is constructed, the historical additive weight and historical waste type in the sample training set are used as input data of the regression network, the historical waste weight in the sample training set is used as output data of the regression network, the regression network is trained to obtain an initial regression network for predicting real-time waste weight, the initial regression network is tested using the sample test set, and a regression network that satisfies a preset error value is output as a second machine learning model, and the regression network is preferably a neural network model.

[0070] It is understandable that the waste weight is obtained through experiments conducted by personnel in this field. Based on the above, the initial additive type and the additive weight correspond to each other, so the additive weight and the waste type will affect the waste weight.

[0071] In this embodiment, in step S20, the weight of the base material and the initial additive type are input into the machine learning model to obtain the weight of the additive. This step ensures that the amount of the additive is accurately controlled according to the different uses of the base material (such as high pressure resistance, infiltration or low sintering uses). Next, step S30 uses the obtained additive weight and initial additive type to further determine the waste type through the machine learning model and predict the waste weight. This combination realizes continuous input and real-time adjustment of data, which not only reduces the waste of raw materials, but also can adjust the production process in time based on the accurate prediction of the weight of each stage, thereby improving the overall efficiency and product quality.

[0072] And through the combination of steps S20 and S30, the entire production process chain is seamlessly connected, especially in scenarios involving different types of base material uses. This combination can ensure that each step from raw material input, additive determination to waste treatment can be coordinated with each other. Specifically, step S20 intelligently selects and quantifies additives according to base material use and base material weight through the first machine learning model. This step directly affects the subsequent waste treatment. When S30 reconfirms the type of waste based on the additive weight and the initial additive type, it can accurately predict the waste weight. In this way, the waste weight can be controlled according to actual needs, thereby reducing waste and optimizing resource recycling. This coordination can avoid unnecessary overhead caused by data isolation in the production process, making the overall process more efficient and stable.

[0073] S40: If the initial additive type is chromite or pyrolusite, determining chromite or pyrolusite as the target additive type;

[0074] It should be noted that, for example, if the initial additive type is chromite, it means that a high-pressure resistant petroleum proppant needs to be prepared. Therefore, chromite is directly used as the target additive type, and chromite is added to the mixing and molding equipment filled with bauxite for mixing to prepare the corresponding petroleum proppant. The same is true for pyrolusite, because there is no other additive that can replace it well. However, dolomite is different, such as Figure 2 As shown, B plus P refers to a petroleum proppant made by mixing bauxite with pyrolusite, and B plus D refers to a petroleum proppant made by mixing dolomite with bauxite. It can be clearly concluded that the sintering temperature of the petroleum proppant mixed with bauxite is about 1350 degrees, while the sintering temperature of the petroleum proppant mixed with pyrolusite is about 1380 degrees, which means that pyrolusite as an additive can also reduce the sintering temperature of the petroleum proppant, so dolomite is replaceable. Since the price of dolomite is lower than that of pyrolusite, when the base material is used for low sintering purposes, dolomite is preferred. However, the price of dolomite and the price of pyrolusite are not constant, and bauxite mixed with dolomite will have burnt materials after processing. The burnt materials need to be treated according to the requirements of environmental protection policies, so there will be costs for treating the burnt materials, and comprehensive considerations are required to determine the target additive type.

[0075] S50: If the initial additive type is dolomite, the weight of the burnt-out material is calculated according to the weight of the additive corresponding to the dolomite and the weight of the waste, the first processing cost is calculated according to the weight of the additive corresponding to the dolomite and the weight of the burnt-out material, the second processing cost is calculated according to the weight of the additive corresponding to the pyrolusite, and the target additive type is determined based on the first processing cost and the second processing cost;

[0076] Specifically, the method for calculating the weight of the burnt material according to the weight of the additive corresponding to the dolomite and the weight of the waste material includes:

[0077] ;

[0078] Where Wfl is the weight of the burned objects, is the weight of additive corresponding to dolomite, is the waste weight corresponding to dolomite, are all weight coefficients.

[0079] It should be noted that bauxite mixed with only dolomite will produce a large amount of burnt materials after processing, while the burnt materials generated by bauxite mixed with pyrolusite or chromite after processing can be ignored. Therefore, when calculating the weight of the burnt materials, it is calculated based on the weight of the additives corresponding to the dolomite and the weight of the waste. From the above content, it can be obtained that the initial additive type, additive weight and waste weight are all corresponding.

[0080] The method for calculating the first processing cost according to the weight of additives corresponding to dolomite and the weight of the burned objects includes:

[0081] ;

[0082] Where IPC is the first processing cost, is the price per unit weight of dolomite, The price corresponding to the unit weight of burned items.

[0083] The method for calculating the second processing cost according to the weight of the additive corresponding to the pyrolusite includes:

[0084] ;

[0085] Where SPC is the second processing cost, is the weight of additive corresponding to pyrolusite, It is the price corresponding to the unit weight of pyrolusite.

[0086] In this embodiment, the weight of additives and waste materials are obtained through the above steps based on the actual production situation and the prediction of the machine learning model. The data obtained in this way not only takes into account the weight of the base material and the use of the base material, but also combines historical data and the learning results of the machine learning model, so that the weight of additives and waste materials calculated each time is more accurate and consistent, which can adapt to slight changes and fluctuations in production, make the first processing cost calculation more accurate and effectively control the volatility of costs, and help enterprises better predict and manage costs.

[0087] The method for determining a target additive type based on a first processing cost and a second processing cost includes:

[0088] When the first processing cost is less than the second processing cost, the dolomite corresponding to the first processing cost is used as the target additive type;

[0089] When the first processing cost is greater than or equal to the second processing cost, the pyrolusite corresponding to the second processing cost is used as the target additive type.

[0090] It is worth noting that, from the above content, it can be concluded that bauxite mixed with dolomite alone will have a large amount of burnt materials after processing, while the burnt materials generated by bauxite mixed with pyrolusite or chromite after processing can be ignored. Therefore, when calculating the cost, only the first processing cost corresponding to dolomite includes the cost of handling the burnt materials. , and Can be directly obtained from the background management system. , and When large fluctuations occur, they can also be modified again by technicians in this field.

[0091] In this embodiment, the weight of the base material and the purpose of the base material are first obtained, the initial additive type is determined according to the purpose of the base material and the first preset rule, the additive weight is obtained according to the weight of the base material and the initial additive type, the waste type is determined according to the initial additive type and the second preset rule, and then the waste weight is obtained through the additive weight and the waste type. If the initial additive type is chromite or pyrolusite, chromite or pyrolusite is determined as the target additive type. If the initial additive type is dolomite, the target additive type is determined based on the first processing cost and the second processing cost. Additives are selected according to demand or purpose, thereby improving the processing quality of the proppant.

[0092] Example 2

[0093] This embodiment provides a petroleum proppant particle forming control system based on machine learning on the basis of embodiment 1, including:

[0094] Data acquisition module: used to obtain the weight of the base material and the purpose of the base material, and determine the initial additive type according to the purpose of the base material and the first preset rule, the initial additive type including chromite, dolomite and pyrolusite;

[0095] It should be noted that the base material in this embodiment refers to the main material for preparing petroleum proppants, and the base material is preferably bauxite. The main components of bauxite are aluminum oxide and silicon dioxide. The weight of the base material can be directly obtained by the sensor in the mixing and molding equipment. The mixing and molding equipment is preferably a horizontal ribbon mixer. It can also be obtained by other means. This embodiment does not impose too many restrictions on this.

[0096] The base material uses include high pressure resistance uses, penetration uses, and low sintering uses. The method for determining the initial additive type according to the base material uses and the first preset rule includes:

[0097] When the base material is used for high pressure resistance, the initial additive type is chromite;

[0098] When the base material is used for penetration, the initial additive type is pyrolusite;

[0099] When the base material application is low sintering application, the initial additive type is dolomite.

[0100] It can be understood that high pressure resistance means that the oil proppant is used in a high pressure environment. Generally, the deeper the formation depth, the greater the pressure generated by the formation fracture, which means that the oil proppant has the advantage of low breakage rate under high pressure environment. Figure 2 To illustrate with an example, Figure 2 B is bauxite, P is pyrolusite, D is dolomite, and C is chromite. Figure 2 It is clear that B plus C refers to the petroleum proppant made of bauxite mixed with chromite, which has the lowest crushing rate. The penetration purpose refers to the formation of a good pore structure inside the finished petroleum proppant, so that oil and gas can flow quickly. B plus P refers to the petroleum proppant made of bauxite mixed with pyrolusite, which has mostly uniformly distributed nearly spherical closed pores inside, so that oil and gas can flow quickly. B plus D refers to the petroleum proppant made of bauxite mixed with dolomite, such as Figure 2 As shown, normally, the sintering temperature of bauxite without additives is about 1480 degrees, but the sintering temperature of bauxite with dolomite added is reduced to about 1350 degrees, so the low sintering application refers to the ability to be processed with lower processing difficulty.

[0101] An additive weight determination module: used to input the base material weight and the initial additive type into a pre-built first machine learning model to obtain the additive weight;

[0102] The first machine learning model construction method includes:

[0103] A sample data set is obtained, wherein the sample data set includes a historical base material weight, a historical initial additive type, and a historical additive weight. The sample data set is divided into a sample training set and a sample test set, and a regression network is constructed. The historical base material weight and the historical initial additive type in the sample training set are used as input data of the regression network, and the historical additive weight in the sample training set is used as output data of the regression network. The regression network is trained to obtain an initial regression network for predicting real-time additive weight. The initial regression network is tested using the sample test set, and a regression network that satisfies a preset error value is output as a first machine learning model. The regression network is preferably a neural network model.

[0104] It is understandable that the weight of the additive is obtained through experiments conducted by personnel in this field. When the weight of the base material is the same, different initial additive types will lead to different additive weights. Therefore, the initial additive type and the additive weight correspond to each other.

[0105] A waste weight determination module: used to determine the type of waste according to the initial additive type and a second preset rule, and input the additive weight and the type of waste into a pre-built second machine learning model to obtain the weight of the waste;

[0106] Specifically, the waste types include calcium oxide and ferric oxide. The waste used in this embodiment is steel slag. The steel slag is screened to extract the main components of calcium oxide and ferric oxide in the steel slag. The main source of calcium oxide with a high content in the steel slag is quicklime added when adjusting the alkalinity during steelmaking. It can be understood that since chromite and pyrolusite both have ferric oxide as one of their main components, and dolomite has calcium oxide as one of its main components, by adding ferric oxide or calcium oxide, chromite, pyrolusite or dolomite can be replaced to a certain extent, which not only recycles the waste, but also reduces the cost of petroleum proppant processing.

[0107] The method for determining the type of waste material according to the initial additive type and the second preset rule includes:

[0108] When the initial additive type is chromite or pyrolusite, the waste type is ferric oxide;

[0109] When the initial additive type is dolomite, the waste type is calcium oxide.

[0110] The second machine learning model construction method includes:

[0111] A sample data set is obtained, wherein the sample data set includes historical additive weight, historical waste type and historical waste weight, the sample data set is divided into a sample training set and a sample test set, a regression network is constructed, the historical additive weight and historical waste type in the sample training set are used as input data of the regression network, the historical waste weight in the sample training set is used as output data of the regression network, the regression network is trained to obtain an initial regression network for predicting real-time waste weight, the initial regression network is tested using the sample test set, and a regression network that satisfies a preset error value is output as a second machine learning model, and the regression network is preferably a neural network model.

[0112] It is understandable that the waste weight is obtained through experiments conducted by personnel in this field. Based on the above, the initial additive type and the additive weight correspond to each other, so the additive weight and the waste type will affect the waste weight.

[0113] The first type determination module is used to determine chromite or pyrolusite as the target additive type if the initial additive type is chromite or pyrolusite;

[0114] It should be noted that, for example, if the initial additive type is chromite, it means that a high-pressure resistant petroleum proppant needs to be prepared. Therefore, chromite is directly used as the target additive type, and chromite is added to the mixing and molding equipment filled with bauxite for mixing to prepare the corresponding petroleum proppant. The same is true for pyrolusite, because there is no other additive that can replace it well. However, dolomite is different, such as Figure 2 As shown, B plus P refers to a petroleum proppant made by mixing bauxite with pyrolusite, and B plus D refers to a petroleum proppant made by mixing dolomite with bauxite. It can be clearly concluded that the sintering temperature of the petroleum proppant mixed with bauxite is about 1350 degrees, while the sintering temperature of the petroleum proppant mixed with pyrolusite is about 1380 degrees, which means that pyrolusite as an additive can also reduce the sintering temperature of the petroleum proppant, so dolomite is replaceable. Since the price of dolomite is lower than that of pyrolusite, when the base material is used for low sintering purposes, dolomite is preferred. However, the price of dolomite and the price of pyrolusite are not constant, and bauxite mixed with dolomite will have burnt materials after processing. The burnt materials need to be treated according to the requirements of environmental protection policies, so there will be costs for treating the burnt materials, and comprehensive considerations are required to determine the target additive type.

[0115] The second type determination module is used for calculating the weight of the burnt-out material according to the weight of the additive corresponding to the dolomite and the weight of the waste material if the initial additive type is dolomite, calculating the first processing cost according to the weight of the additive corresponding to the dolomite and the weight of the burnt-out material, calculating the second processing cost according to the weight of the additive corresponding to the pyrolusite, and determining the target additive type based on the first processing cost and the second processing cost;

[0116] Specifically, the method for calculating the weight of the burnt material according to the weight of the additive corresponding to the dolomite and the weight of the waste material includes:

[0117] ;

[0118] Where Wfl is the weight of the burned objects, is the weight of additive corresponding to dolomite, is the waste weight corresponding to dolomite, are all weight coefficients.

[0119] It should be noted that bauxite mixed with only dolomite will produce a large amount of burnt materials after processing, while the burnt materials generated by bauxite mixed with pyrolusite or chromite after processing can be ignored. Therefore, when calculating the weight of the burnt materials, it is calculated based on the weight of the additives corresponding to the dolomite and the weight of the waste. From the above content, it can be obtained that the initial additive type, additive weight and waste weight are all corresponding.

[0120] The method for calculating the first processing cost according to the weight of additives corresponding to dolomite and the weight of the burned objects includes:

[0121] ;

[0122] Where IPC is the first processing cost, is the price per unit weight of dolomite, The price corresponding to the unit weight of burned items.

[0123] The method for calculating the second processing cost according to the weight of the additive corresponding to the pyrolusite includes:

[0124] ;

[0125] Where SPC is the second processing cost, is the weight of additive corresponding to pyrolusite, It is the price corresponding to the unit weight of pyrolusite.

[0126] The method for determining a target additive type based on a first processing cost and a second processing cost includes:

[0127] When the first processing cost is less than the second processing cost, the dolomite corresponding to the first processing cost is used as the target additive type;

[0128] When the first processing cost is greater than or equal to the second processing cost, the pyrolusite corresponding to the second processing cost is used as the target additive type.

[0129] It is worth noting that, from the above content, it can be concluded that bauxite mixed with dolomite alone will have a large amount of burnt materials after processing, while the burnt materials generated by bauxite mixed with pyrolusite or chromite after processing can be ignored. Therefore, when calculating the cost, only the first processing cost corresponding to dolomite includes the cost of handling the burnt materials. , and Can be directly obtained from the background management system. , and When large fluctuations occur, they can also be modified again by technicians in this field.

[0130] Example 3

[0131] This embodiment discloses an electronic device, including a power supply, an interface, a keyboard, a memory, a central processing unit, and a computer program stored in the memory and executable on the central processing unit. When the central processing unit executes the computer program, the machine learning-based petroleum proppant particle forming control method provided by the above-mentioned methods is implemented. The interface includes a network interface and a data interface. The network interface includes a wired or wireless interface, and the data interface includes an input or output interface.

[0132] Since the electronic device introduced in this embodiment is an electronic device used to implement the petroleum proppant particle forming control method based on machine learning in the embodiment of this application, based on the petroleum proppant particle forming control method based on machine learning introduced in the embodiment of this application, the technical personnel of this field can understand the specific implementation of the electronic device of this embodiment and its various variations, so how the electronic device implements the method in the embodiment of this application is not described in detail here. As long as the technical personnel of this field implement the electronic device used by the petroleum proppant particle forming control method based on machine learning in the embodiment of this application, it belongs to the scope of protection of this application.

[0133] Example 4

[0134] like Figure 3 As shown, this embodiment discloses a computer-readable storage medium, on which a computer program is stored. When the computer program is executed, the above-mentioned machine learning-based petroleum proppant particle forming control method is implemented.

[0135] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters, weights and thresholds in the formula are set by technicians in this field according to actual conditions.

[0136] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website, computer, server or data center to another website, computer, server or data center via a wired network or a wireless network. The computer-readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server or data center that includes one or more available media sets. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD) or a semiconductor medium. The semiconductor medium may be a solid-state hard disk.

[0137] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the present invention can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0138] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0139] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the units is only one, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0140] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0141] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0142] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

[0143] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for controlling the formation of petroleum proppant particles based on machine learning, characterized in that: include: S10: Obtaining the weight of the base material and the purpose of the base material, and determining the type of initial additive according to the purpose of the base material and the first preset rule, the initial additive type including chromite, dolomite and pyrolusite, and the base material is bauxite; S20: Inputting the weight of the base material and the initial additive type into a pre-built first machine learning model to obtain the weight of the additive; S30: Determine the type of waste material according to the initial additive type and the second preset rule, input the weight of the additive and the type of waste material into a pre-built second machine learning model, and obtain the weight of the waste material; S40: If the initial additive type is chromite or pyrolusite, determining chromite or pyrolusite as the target additive type; S50: If the initial additive type is dolomite, the weight of the burnt-out material is calculated according to the weight of the additive corresponding to the dolomite and the weight of the waste, the first processing cost is calculated according to the weight of the additive corresponding to the dolomite and the weight of the burnt-out material, the second processing cost is calculated according to the weight of the additive corresponding to the pyrolusite, and the target additive type is determined based on the first processing cost and the second processing cost; The method for calculating the weight of the burnt material according to the weight of the additive corresponding to the dolomite and the weight of the waste material includes: ; Where Wfl is the weight of the burned objects, is the weight of additive corresponding to dolomite, is the waste weight corresponding to dolomite, and All are weight coefficients; The method for calculating the first processing cost according to the weight of the additive corresponding to the dolomite and the weight of the burned objects comprises: ; Where IPC is the first processing cost, is the price per unit weight of dolomite, The price corresponding to the unit weight of burned items.

2. The method for controlling the formation of petroleum proppant particles based on machine learning according to claim 1, characterized in that: The base material uses include high pressure resistance, penetration and low sintering. The method for determining the initial additive type according to the base material use and the first preset rule includes: When the base material is used for high pressure resistance, the initial additive type is chromite; When the base material is used for penetration, the initial additive type is pyrolusite; When the base material application is low sintering application, the initial additive type is dolomite.

3. The method for controlling the formation of petroleum proppant particles based on machine learning according to claim 2, characterized in that: The method for constructing the first machine learning model includes: A sample data set is obtained, wherein the sample data set includes a historical base material weight, a historical initial additive type, and a historical additive weight. The sample data set is divided into a sample training set and a sample test set, and a regression network is constructed. The historical base material weight and the historical initial additive type in the sample training set are used as input data of the regression network, and the historical additive weight in the sample training set is used as output data of the regression network. The regression network is trained to obtain an initial regression network for predicting real-time additive weight. The initial regression network is tested using the sample test set, and the regression network that satisfies a preset error value is output as a first machine learning model.

4. The method for controlling the formation of petroleum proppant particles based on machine learning according to claim 1, characterized in that: The waste types include calcium oxide and ferric oxide, and the method for determining the waste type according to the initial additive type and the second preset rule includes: When the initial additive type is chromite or pyrolusite, the waste type is ferric oxide; When the initial additive type is dolomite, the waste type is calcium oxide.

5. The method for controlling the formation of petroleum proppant particles based on machine learning according to claim 4, characterized in that: The method for constructing the second machine learning model includes: A sample data set is obtained, wherein the sample data set includes historical additive weights, historical waste types, and historical waste weights; the sample data set is divided into a sample training set and a sample test set; a regression network is constructed; the historical additive weights and historical waste types in the sample training set are used as input data of the regression network; the historical waste weights in the sample training set are used as output data of the regression network; the regression network is trained to obtain an initial regression network for predicting real-time waste weight; the initial regression network is tested using the sample test set; and a regression network that satisfies an error value less than a preset value is output as a second machine learning model.

6. The method for controlling the formation of petroleum proppant particles based on machine learning according to claim 1, characterized in that: The method for calculating the second processing cost according to the weight of the additive corresponding to the pyrolusite comprises: ; Where SPC is the second processing cost, is the weight of additive corresponding to pyrolusite, It is the price corresponding to the unit weight of pyrolusite.

7. The method for controlling the formation of petroleum proppant particles based on machine learning according to claim 1, characterized in that: The method for determining the target additive type based on the first processing cost and the second processing cost includes: When the first processing cost is less than the second processing cost, the dolomite corresponding to the first processing cost is used as the target additive type; When the first processing cost is greater than or equal to the second processing cost, the pyrolusite corresponding to the second processing cost is used as the target additive type.

8. A machine learning-based petroleum proppant particle forming control system, which is used to implement the machine learning-based petroleum proppant particle forming control method according to any one of claims 1 to 7, characterized in that: include: Data acquisition module: used to obtain the weight of the base material and the purpose of the base material, and determine the initial additive type according to the purpose of the base material and the first preset rule, the initial additive type including chromite, dolomite and pyrolusite; An additive weight determination module: used to input the base material weight and the initial additive type into a pre-built first machine learning model to obtain the additive weight; A waste weight determination module: used to determine the type of waste according to the initial additive type and a second preset rule, and input the additive weight and the type of waste into a pre-built second machine learning model to obtain the weight of the waste; The first type determination module is used to determine chromite or pyrolusite as the target additive type if the initial additive type is chromite or pyrolusite; The second type determination module is used to calculate the weight of the burnt material according to the weight of the additive corresponding to the dolomite and the weight of the waste if the initial additive type is dolomite, calculate the first processing cost according to the weight of the additive corresponding to the dolomite and the weight of the burnt material, calculate the second processing cost according to the weight of the additive corresponding to the pyrolusite, and determine the target additive type based on the first processing cost and the second processing cost.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed, the method for controlling the formation of petroleum proppant particles based on machine learning as described in any one of claims 1 to 7 is implemented.

Images