Well drilling big data optimization method for increasing speed and efficiency

Abstract

The invention provides a drilling big data optimization method for increasing speed and efficiency, which is based on a big data processing thought, comprehensively considers rock mechanical parameters and a drill bit use effect, and establishes an adaptability coefficient weighted calculation equation conforming to drill bit characteristics. Secondly, on the basis of a drill bit optimization result, composite drilling analysis of hydraulic jet and drill bit mechanical specific energy is carried out, and finally, a drilling big data real-time parameter iterative optimization model is established in combination with a particle swarm algorithm and properly assisted by neural network analysis and iterative verification, so that on-site huge real-time data can be well matched, the mechanical drilling speed can be increased, and the drilling efficiency can be improved. Accidents in actual working conditions are reduced, well drilling parameters can be conveniently adjusted on site in real time, well drilling risks are effectively avoided, and well drilling efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of oil drilling, and in particular to a drilling big data optimization method for increasing speed and efficiency. Background Art

[0002] Mechanical drilling speed is a key factor affecting the drilling operation cycle and operating cost, which can directly reflect the drilling efficiency and drilling operation level. In this regard, domestic and foreign experts and scholars have established many speed-up optimization models based on the drilling system. However, in the face of the increasingly complex drilling operation environment and massive field data, the existing speed-up optimization models cannot match the huge real-time data on site well, the solution process is complex, and the calculation accuracy is limited.

[0003] In recent years, artificial intelligence and big data technologies have been applied in many fields, providing new technical ideas for optimizing drilling parameters and improving mechanical drilling speed. How to use big data analysis to optimize drilling speed in real time, reduce the complexity of drilling accidents, and improve mechanical drilling speed is of great significance to the digital development of oil fields. Summary of the invention

[0004] The present invention aims to provide a drilling big data optimization method for increasing speed and efficiency, so as to solve the problem that the existing speed-up optimization model cannot well match the huge real-time data on site.

[0005] The present invention provides a drilling big data optimization method for improving speed and efficiency, which mainly optimizes the following aspects:

[0006] (1) Determine the mechanical specific energy and match the optimal drill bit with the optimal drilling speed;

[0007] (2) Based on particle swarm optimization, determine whether the current parameters are optimal;

[0008] (3) Real-time optimization model, which optimizes the non-optimal drilling parameters in real time and obtains the drilling parameter optimization results.

[0009] The drilling big data optimization method for increasing speed and efficiency comprises the following steps:

[0010] S1. Quantify the structural characteristics according to the drill bit model and optimize the drill bit big data based on the golden section method;

[0011] S2. Based on the optimization results of drill bit big data, conduct composite drilling analysis of hydraulic jet and drill bit mechanical specific energy;

[0012] S3. Combine the composite drilling analysis results with the particle swarm optimization algorithm to establish a drilling big data real-time parameter iterative optimization model; use the drilling big data real-time parameter iterative optimization model to optimize the non-optimal drilling parameters in real time to obtain the drilling parameter optimization results.

[0013] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0014] The present invention is based on the idea of ​​big data processing, comprehensively considers the rock mechanical parameters and the use effect of the drill bit, and establishes a weighted calculation equation for the fitness coefficient that meets the characteristics of the drill bit. Secondly, based on the drill bit optimization results, a composite drilling analysis of hydraulic jet and drill bit mechanical specific energy is carried out. Finally, combined with the particle swarm algorithm, appropriately supplemented by neural network analysis and iterative verification, a drilling big data real-time parameter iterative optimization model is established, which can well match the huge real-time data on site, can improve the mechanical drilling speed, and reduce accidents in actual working conditions. It has a good effect, is convenient for real-time adjustment of drilling parameters on site, effectively avoids drilling risks, and improves drilling efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0016] Figure 1 The present invention is a flowchart of a drilling big data optimization method for increasing speed and efficiency in an embodiment of the present invention.

[0017] Figure 2 Schematic diagram of the preferred drill bit for the golden section method in an embodiment of the present invention.

[0018] Figure 3 Schematic diagram of the relationship between mechanical specific energy and mechanical drilling speed in an embodiment of the present invention.

[0019] Figure 4 Modeling flow chart for the real-time optimization model of drilling speed based on particle swarm optimization algorithm. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0021] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. 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.

[0022] Example

[0023] like Figure 1 As shown, this embodiment proposes a drilling big data optimization method for increasing speed and efficiency, comprising the following steps:

[0024] S1. Quantify the structural characteristics according to the drill bit model and optimize the drill bit big data based on the golden section method;

[0025] Traditional drill bit optimization methods often rely on rock mechanics parameters, but due to the limitations of rock mechanics analysis data and testing methods, the drill bit optimization results are not sufficient to fully adapt to the engineering conditions of the formation. Therefore, starting from big data thinking, combined with real-time drilling parameters, comprehensive consideration of key rock mechanics parameters, and multi-dimensional quantitative description of drill bit characteristics can make up for the defect of poor drill bit optimization effect due to fewer parameters. According to the structural characteristics of the drill bit, a drill bit big data optimization method is formed, such as Figure 1 Step S1 includes the following sub-steps:

[0026] S11, quantification of structural characteristics by drill bit model:

[0027] (1) Construct a rock mechanics parameter evaluation template based on logging data, establish a uniaxial compressive strength (UCS) profile curve, and clarify the uniaxial compressive strength range of different layers through big data analysis;

[0028] (2) Quantify the structural characteristic parameters of the PDC drill bit in each uniaxial compressive strength range, as shown in Table 1.

[0029] Table 1, Quantitative table of structural characteristic parameters of PDC drill bits:

[0030]

[0031]

[0032] (3) Determine the applicability coefficient of the uniaxial compressive strength range and the structural characteristic parameters of the PDC drill bit;

[0033] For example, the PDC drill bit tooth density characteristic code list is explained as shown in Table 2.

[0034] Table 2, PDC drill bit tooth density characteristic code:

[0035] Density code 1 2 3 4 Density level dilute middle dense high density Quantity range ≤30 ＞30～40 ＞40～50 ＞50

[0036] The equivalent number of teeth of PDC drill bits is divided based on "Φ215.9mm diameter PDC drill bit + Φ13.44mm tooth shape". If the actual PDC drill bit is different from this benchmark, it should be converted. The conversion method is:

[0037]

[0038] Where:

[0039] Z is the equivalent number of teeth of the PDC drill bit, dimensionless;

[0040] D is the diameter of the PDC drill bit, mm;

[0041] n is the number of cutting teeth types on the PDC drill bit (cutting teeth with the same diameter are classified into one category), dimensionless;

[0042] Z i is the number of the i-th teeth on the PDC drill bit, dimensionless;

[0043] d i is the diameter of the i-th tooth on the PDC drill bit, mm.

[0044] According to the height range of the crown of the PDC drill bit, it is divided into short crown (height less than 58mm), medium crown (height between 58 and 114mm), and long crown (height greater than 114mm). The calculation formula for the applicable coefficient of the PDC drill bit is expressed as:

[0045]

[0046] Where: x is the uniaxial compressive strength, MPa;

[0047] m i , b are parameters to be determined, and i is the structural characteristic parameter code of the PDC drill bit.

[0048] Among them, the comparative relationship between the uniaxial compressive strength range and the structural characteristic parameters of the PDC drill bit is shown in Table 3.

[0049] Table 3, PDC drill bit cutting structure characteristic parameter comparison table:

[0050]

[0051] (4) Within the range of the formation section drilled in one well, the applicable coefficient of the structural characteristic parameters of the PDC drill bit within each uniaxial compressive strength range is weighted by the number of footage, and the calculation formula is expressed as follows:

[0052]

[0053] Where:

[0054] F(x i ) is a calculation function for the weighted value of the applicable coefficient of the structural characteristic parameters of the PDC drill bit within the range of uniaxial compressive strength;

[0055] K is the total uniaxial compressive strength;

[0056] j is the range number of the uniaxial compressive strength interval;

[0057] f j (x i ) represents the applicability coefficient of the structural characteristic parameters of the ith PDC drill bit within the jth uniaxial compressive strength range.

[0058] The above formula obtains the applicable coefficient weighted value of the PDC drill bit structural characteristic parameter within the uniaxial compressive strength range. The PDC drill bit structural characteristic parameter corresponding to the maximum applicable coefficient weighted value is the most suitable PDC drill bit within the uniaxial compressive strength range.

[0059] S12, a method for optimizing drill bit big data based on the golden section method, such as Figure 2 As shown, specifically including:

[0060] (1) Determine the golden section value h:

[0061]

[0062] (2) Determine any point X on the golden section line n :

[0063] X n =nV max / m

[0064] (3) Determine any point Y on the golden section line n :

[0065] Y n =h / X n

[0066] Where:

[0067] is the overall average mechanical drilling speed of the target formation, m / h;

[0068] The average single drill bit penetration of the target formation, m / bit;

[0069] V max The highest average mechanical drilling speed of a single drill bit in the target formation, m / h;

[0070] m is the number of golden section points, and it is sufficient to ensure that the golden section line is smooth;

[0071] X n , Y n is the X value and Y value of the nth point of the golden section line.

[0072] (4) Based on big data analysis, the use effect of PDC drill bits in different layers is evaluated, and the PDC drill bits with good use effect are selected as basic samples using the golden section method, and the applicability coefficients of PDC drill bit structural characteristic parameters in different uniaxial compressive strength ranges are revised; according to the uniaxial compressive strength range of the section to be drilled, the revised applicability coefficients are used to determine the preferred PDC drill bit structural characteristic parameters, match the manufacturer and drill bit model, and determine the PDC drill bit big data optimization results. The PDC drill bit big data optimization results are applied on site, and the application data are added to the database in real time, and the method for drill bit big data optimization based on the golden section method is updated and improved.

[0073] S2. Based on the optimization results of drill bit big data, conduct composite drilling analysis of hydraulic jet and drill bit mechanical specific energy;

[0074] (1) The ideal mechanical specific energy value is equal to the rock compressive strength. The actual drilling is affected by many factors, resulting in the actual mechanical specific energy being equal to about 3 times the rock compressive strength. Figure 3 As shown in the figure, mechanical specific energy is defined as the work done by the drill bit to break a unit volume of rock under the action of drilling pressure and torque. The calculation formula is expressed as:

[0075]

[0076] Where:

[0077] MSE is mechanical specific energy, MPa;

[0078] W is the drilling pressure, N;

[0079] A b The cross-sectional area drilled by the drill bit, mm 2 ;

[0080] T is torque, N·m;

[0081] λ is the rotation speed, r / min;

[0082] ν is the mechanical drilling speed, m / h.

[0083] (2) The correction coefficient is introduced to correct the mechanical specific energy so that the mechanical specific energy is closer to the true strength of the rock. The calculation formula of the corrected mechanical specific energy is expressed as:

[0084] MSE m =E f MSE

[0085] Where:

[0086] MSE m is the corrected mechanical specific energy, MPa;

[0087] E f is the correction factor, dimensionless.

[0088] (3) The drill bit sliding friction coefficient is introduced, the torque calculation method is optimized, and the drill bit wear is analyzed through the mechanical specific energy trend line, making the drill bit selection, fault diagnosis and drilling efficiency evaluation more accurate. The calculation formula is expressed as:

[0089]

[0090]

[0091] Where:

[0092] D b is the drill bit diameter, mm;

[0093] μ is the drill bit sliding friction coefficient, dimensionless.

[0094] (4) A multivariate linear regression model of drill bit torque is established. The parameters required for torque calculation can be obtained through ground instruments to improve the applicability of the original mechanical specific energy model. The calculation model is as follows:

[0095]

[0096] (5) Based on the multivariate linear regression model of the drill bit torque, the rock breaking and clearing effect of hydraulic jet is considered. When the mechanical energy and hydraulic energy increase at the same time, the drilling specific energy decreases, which effectively improves the mechanical drilling speed. The calculation principle is as follows:

[0097]

[0098]

[0099] Where: DSE is the improved mechanical specific energy, MPa;

[0100] λ W is the hydraulic correlation coefficient, dimensionless; HP B is the drill water power, W.

[0101] (6) Composite drilling refers to adding downhole power drill tools to the downhole drill tool combination to convert the hydraulic energy of the drilling fluid circulation into mechanical energy of the drill bit, thereby breaking the rock and drilling. The rotation speed can be adjusted by the drilling fluid flow rate. In the composite drilling process, the drill bit drive consists of a surface drive (rotary table or top drive) and an underground drive (screw drill), of which the underground drive is the main power of the drill bit. Considering the mechanical specific energy and hydraulic jet, the main performance parameters of the screw power drill are torque and speed. Its theoretical speed is only related to the drill flow rate and the displacement per revolution of the drill, that is:

[0102]

[0103] Where: q is the displacement per revolution of the drill, L / r; K N is the speed-flow ratio of the power drilling tool, r / L.

[0104] S3. Based on the composite drilling analysis combined with the particle swarm optimization algorithm, a real-time parameter iteration optimization model for drilling big data is established; the real-time parameter iteration optimization model for drilling big data is used to optimize the non-optimal drilling parameters in real time to obtain the drilling parameter optimization results.

[0105] According to the mechanical specific energy of composite drilling analysis, the optimization effect is analyzed with the actual drilling mechanical specific energy curve, and the main drilling parameters that affect the drilling speed are determined. The curve of the highest value that rock breaking efficiency can reach during the drilling process is defined as the mechanical specific energy base value line, which represents the overall change of rock strength in a certain formation. The deviation of the actual mechanical specific energy curve from the mechanical specific energy base value line reflects the change trend of the formation physical properties of the block under the same drilling parameters.

[0106] The optimization effect is analyzed by the actual drilling mechanical specific energy curve. If the rock strength baseline value is significantly smaller than the actual mechanical specific energy value, the rock breaking efficiency is low. The drilling parameters need to be adjusted according to the downhole working conditions to improve the rock breaking efficiency of the drill bit. The process is as follows:

[0107] (1) When the actual mechanical specific energy value deviates greatly from the mechanical specific energy baseline value, adjust the drilling pressure first, then adjust the speed, and the adjustment range should not exceed 10%;

[0108] (2) When adjusting the drilling pressure, first reduce the drilling pressure. After drilling for a period of time, if the actual mechanical specific energy value still deviates from the mechanical specific energy baseline value, restore the drilling pressure and then increase the drilling pressure to determine the mechanical specific energy value;

[0109] (3) If adjusting the drilling pressure does not improve the rock breaking efficiency, restore the drilling pressure and adjust the speed by first decreasing and then increasing it;

[0110] (4) When the displacement is constant, if adjusting the drilling pressure and speed is ineffective and the actual mechanical specific energy value still increases, it means that the drill bit is not adapted to the formation or the drill bit is severely worn and needs to be replaced.

[0111] The particle swarm optimization (PSO) algorithm is introduced to find the optimal solution for drilling parameters. Assuming that there is only one optimal solution in the region, each particle searches for the optimal solution in the search space individually and records it as the current individual extreme value. The individual extreme value is shared with other particles in the entire particle swarm to determine whether the optimal solution is found.

[0112] The optimal solution update formula is expressed as:

[0113]

[0114] Where:

[0115] ω is the inertia factor;

[0116] R is a random number whose absolute value does not exceed 1;

[0117] C1 and C2 are acceleration constants, generally C1=C2∈[0,4];

[0118] P id is the dth dimension of the individual extreme value of the th variable;

[0119] P gd is the dth dimension of the global optimal solution.

[0120] Calculate the covariance to measure the overall error of the variables. If the change trends are consistent, the covariance is positive; if the change trends are opposite, the covariance is negative; if the variables are independent of each other, the covariance is 0, that is, the two variables are unrelated. The covariance calculation formula for variable parameters X and Y is:

[0121]

[0122] The correlation of different drilling parameters is calculated to obtain the correlation matrix. The drilling parameters with the highest correlation are obtained through correlation matrix analysis and used as model input parameters, including displacement, rotation speed, drilling pressure, drilling fluid density, funnel viscosity, formation density, uniaxial compressive strength and other parameters with actual physical significance. The model input parameters and their corresponding drilling speed are cleaned, and the cleaned data is regularized and normalized. The specific function is:

[0123]

[0124] Where:

[0125] m is the number of samples;

[0126] w is the weight coefficient vector;

[0127] δ is a hyperparameter used to control the degree of regularization;

[0128] b is the training sample and the corresponding label;

[0129] Ω(w) is the sum of the squares of the absolute values ​​of all input parameters;

[0130] The drilling speed at each moment is used as the output parameter (y) of the machine learning algorithm, and the remaining drilling parameters are used as the input parameter set {x} of the machine learning algorithm. The initial predicted drilling speed can be obtained through neural network calculation, which is recorded as y. The training residual is obtained by calculation:

[0131]

[0132] Use the training residual to reverse the input parameter x i The weight value w between the neural network layers i To make corrections:

[0133]

[0134] The weight value w i Corrected to w in :

[0135]

[0136] Where: η is the learning rate; w i is the weight value; w in is the corrected weight value.

[0137] By applying the particle swarm optimization algorithm, the optimal drilling parameters can be obtained through neural network analysis iteration. The displacement, rotation speed and drilling pressure are selected as the controllable parameters for optimizing the drilling speed. The other drilling parameters are used as fixed parameters and input into the optimal drilling parameter model. When the objective function and the number of iterations of the LSTM-PSO algorithm are stable, the real-time optimization of the drilling parameters is finally achieved.

[0138] The drill bit model recommendation and drilling parameter optimization were completed separately, and finally a real-time parameter iterative optimization model for drilling big data was established. The key parameters are shown in Table 4.

[0139] Table 4, drill bit models and drilling parameters for each drilling:

[0140]

[0141] Application example: The drilling big data optimization method of the present invention is reliable in the drilling site of the Yang 101 well area in the Luzhou block, and the mechanical drilling speed is significantly improved. This proves that the method has a good effect on improving the mechanical drilling speed and reducing accidents in actual working conditions, and is convenient for adjusting drilling parameters in real time on site, effectively avoiding drilling risks, and improving drilling efficiency.

[0142] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A drilling big data optimization method to increase speed and efficiency. It is characterized in that The steps include: S1. Quantify the structural characteristics according to the drill bit model and optimize the drill bit big data based on the golden section method; S2. Based on the optimization results of drill bit big data, conduct composite drilling analysis of hydraulic jet and drill bit mechanical specific energy; S3. Combine the composite drilling analysis results with the particle swarm optimization algorithm to establish a drilling big data real-time parameter iterative optimization model; use the drilling big data real-time parameter iterative optimization model to optimize the non-optimal drilling parameters in real time to obtain the drilling parameter optimization results.

2. The drilling big data optimization method for increasing speed and efficiency according to claim 1, It is characterized in that The method for quantifying structural features according to drill bit model in step S1 includes: (1) Construct a rock mechanics parameter evaluation template based on logging data, establish a uniaxial compressive strength profile curve, and clarify the uniaxial compressive strength range of different layers through big data analysis; (2) Quantify the structural characteristic parameters of the drill bit in each uniaxial compressive strength range; (3) Determine the applicability coefficient of the uniaxial compressive strength range and the drill bit structural characteristic parameters; (4) The drill bit structural characteristic parameter applicability coefficient within each uniaxial compressive strength range is weighted by the number of penetrations.

3. According to the drilling big data optimization method for increasing speed and efficiency as described in claim 2, It is characterized in that The calculation formula of the structural characteristic parameter drill bit applicability coefficient is expressed as: Where: x is the uniaxial compressive strength; m i , b are parameters to be determined, and i is the code for the characteristic parameter of the drill bit structure.

4. The drilling big data optimization method for increasing speed and efficiency according to claim 3, It is characterized in that The calculation formula for weighting the applicable coefficient of the drill bit structural characteristic parameters within each uniaxial compressive strength range with the footage number is expressed as: Where: F(x i ) is the calculation function of the weighted value of the applicable coefficient of the drill bit structural characteristic parameters within the range of uniaxial compressive strength; K is the total number of uniaxial compressive strength; j is the range number of uniaxial compressive strength; f j (x i ) represents the applicability coefficient of the i-th drill bit structural characteristic parameter within the j-th uniaxial compressive strength range.

5. The drilling big data optimization method for increasing speed and efficiency according to claim 1, It is characterized in that The method for optimizing drill bit big data based on the golden section method in step S1 includes: (1) Determine the golden section value h: (2) Determine any point X on the golden section line n : X n =nV max / m (3) Determine any point Y on the golden section line n : Y n =h / X n Where: is the overall average mechanical drilling speed of the target formation; V is the average single drill bit penetration of the target formation; max is the average mechanical drilling speed of the highest single drill bit in the target formation; m is the number of golden section points; X n , Y n The X and Y values ​​of the nth point of the golden section line; (4) Based on big data analysis, the use effect of drill bits in different layers is evaluated, and the drill bits with good use effect are selected as basic samples using the golden section method, and the applicability coefficients of the drill bit structural characteristic parameters in different uniaxial compressive strength ranges are revised; according to the uniaxial compressive strength range of the section to be drilled, the preferred drill bit structural characteristic parameters are determined using the revised applicability coefficients, and the manufacturer and drill bit model are matched to determine the drill bit big data optimization results; the drill bit big data optimization results are applied on-site, and the application data are added to the database in real time, and the method for drill bit big data optimization based on the golden section method is updated and improved.

6. The drilling big data optimization method for increasing speed and efficiency according to any one of claims 2 to 5, It is characterized in that Step S2 includes: (1) Mechanical specific energy is defined as the work done by the drill bit to break a unit volume of rock under the action of drilling pressure and torque. The calculation formula is expressed as: Where: MSE is mechanical specific energy; W is drilling pressure; A b is the cross-sectional area drilled by the drill bit; T is the torque; λ is the rotation speed; ν is the mechanical drilling speed; (2) The correction coefficient is introduced to correct the mechanical specific energy so that the mechanical specific energy is closer to the true strength of the rock. The calculation formula of the corrected mechanical specific energy is expressed as: MSE m =E f ·MSE Where: MSE m E is the corrected mechanical specific energy; f is the correction factor; (3) The drill bit sliding friction coefficient is introduced to optimize the torque calculation method. The calculation formula is expressed as: Where: D b is the drill bit diameter; μ is the drill bit sliding friction coefficient; (4) Establish a multivariate linear regression model of drill bit torque, expressed as: (5) Based on the multivariate linear regression model of the drill bit torque, the mechanical specific energy is improved and expressed as: Where: DSE is the improved mechanical specific energy; λ W is the hydraulic correlation coefficient; HP B is the drill water power; (6) The composite drilling analysis of hydraulic jet and drill bit mechanical specific energy is completed through the improved mechanical specific energy, and the mechanical specific energy calculation formula is obtained as follows: Where: q is the displacement per revolution of the drill bit; K N is the speed-flow ratio of the power drilling tool.

7. The drilling big data optimization method for increasing speed and efficiency according to claim 6, It is characterized in that Step S3 includes: (1) According to the mechanical specific energy of composite drilling analysis, the optimization effect is analyzed with the actual drilling mechanical specific energy curve to determine the main drilling parameters that affect the drilling speed; (2) The particle swarm optimization algorithm is introduced to find the optimal solution for drilling parameters. Assuming that there is only one optimal solution in the region, each particle searches for the optimal solution in the search space separately and records it as the current individual extreme value. The individual extreme value is shared with other particles in the entire particle swarm to determine whether the optimal solution is found. The optimal solution update formula is expressed as: Where: ω is the inertia factor; R is a random number whose absolute value does not exceed 1; C 1 , C 2 is the acceleration constant; P id is the dth dimension of the individual extreme value of the dth variable; P gd is the dth dimension of the global optimal solution; (3) Calculate the covariance to measure the overall error of the variables; if the change trends are consistent, the covariance is positive; if the change trends are opposite, the covariance is negative; if the variables are independent of each other, the covariance is 0, that is, the two variables are unrelated; the covariance calculation formula of the variable parameters X and Y in the optimal solution update formula is: (4) Calculate the correlation of different drilling parameters to obtain the correlation matrix. The drilling parameter with the highest correlation is obtained through correlation matrix analysis, and the drilling parameter with the highest correlation is used as the model input parameter. The model input parameter and the corresponding drilling speed are cleaned, and the cleaned data is regularized and normalized, which is expressed as: Where: m is the number of samples; w is the weight coefficient vector; δ is a hyperparameter used to control the degree of regularization; b is the training sample and the corresponding label; Ω(w) is the sum of the squares of the absolute values ​​of all input parameters; (5) The drilling speed at each moment is taken as the output parameter y of the machine learning algorithm, and the remaining engineering parameters are taken as the input parameter set {x} of the machine learning algorithm. The initial predicted drilling speed is obtained by neural network calculation, which is recorded as y. The training residual is obtained by calculation: Use the training residual to reverse the input parameter x i The weight value w between the neural network layers i To make corrections: The weight value w i Corrected to w in : Where: η is the learning rate; w i is the weight value; w in is the corrected weight value; The particle swarm optimization algorithm is applied to obtain the optimal drilling parameters through neural network analysis iteration. The displacement, rotation speed and drilling pressure are selected as the controllable parameters of the optimal drilling speed. The other drilling parameters are used as fixed parameters and input into the optimal drilling parameter model. When the objective function and the number of iterations of the LSTM-PSO algorithm are stable, the real-time optimization of the drilling parameters is finally achieved. (6) The drill bit model recommendation and drilling parameter optimization are completed separately, and finally a real-time parameter iterative optimization model for drilling big data is established.

8. The drilling big data optimization method for increasing speed and efficiency according to claim 7, It is characterized in that C 1 =C 2 ∈[0,4]。

Images