MGO-based logging-while-drilling instrument digital twinborn model drilling parameter prediction method

Abstract

The invention relates to the technical field of logging while drilling, in particular to an MGO-based logging while drilling instrument digital twinborn model drilling parameter prediction method, which comprises the following steps: constructing a motion model of a logging while drilling azimuth electromagnetic wave logging instrument, and obtaining drilling parameters; the unit drilling length cost, the mechanical specific energy and the drilling and measuring speed serve as objective functions, and constraint conditions of the drill bit abrasion loss, the drilling tool upper portion drilling pressure and the instrument rotating speed are set; and optimizing the target function by using an MGO algorithm. According to the method, the problems that an existing meta-heuristic optimization algorithm is uneven in optimal solution distribution and the prediction accuracy needs to be further improved are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of logging while drilling, and particularly to a method for predicting drilling parameters of a digital twin model of a logging while drilling tool based on MGO. Background Art

[0002] The azimuthal electromagnetic wave logging while drilling technology is one of the key irreplaceable technologies in oil exploration and development. It can effectively improve the reservoir drilling encounter rate of highly deviated wells and horizontal wells, providing strong support for oil and gas resource exploration and development.

[0003] The prior art uses the numerical simulation method. Although it can achieve high calculation accuracy, it is difficult to overcome the problems of high model complexity, huge calculation amount, and inability to be applied to real-time geological steering while drilling; moreover, it is also difficult to fully model, interpret, invert, and compensate the effects of fatigue, wear, aging, and faults of key components inside the instrument and the stress of the downhole complex environment on the logging response of the instrument through numerical simulation methods.

[0004] Meta-heuristic optimization algorithms are used for prediction in logging while drilling, such as the SSA algorithm, GWO algorithm, and PSO; Zhang Haibo et al.'s research on optimizing drilling parameters in the Doseo block uses the PSO algorithm for logging while drilling prediction, but there are problems such as uneven distribution of the optimal solution and low accuracy of the optimal solution when the number of iterations is small. Summary of the Invention

[0005] Aiming at the deficiencies of the existing methods, the present invention solves the problems of uneven distribution of the optimal solution and the need for further improvement in prediction accuracy existing in the existing meta-heuristic optimization algorithms.

[0006] The technical solution adopted by the present invention is as follows: The method for predicting drilling parameters of a digital twin model of a logging while drilling tool based on MGO includes the following steps:

[0007] Step 1: Construct a motion model of the azimuthal electromagnetic wave logging while drilling instrument to obtain drilling parameters;

[0008] As a preferred embodiment of the present invention, the drilling parameters include the drilling speed v, the instrument rotation speed ω, the power P of the turbine mud generator, the pitch angle θ, the weight on bit F1 of the upper part of the drill string, and the instrument torque T.

[0009] As a preferred embodiment of the present invention, the formula for the instrument torque is:

[0010]

[0011] T = T f + T w (8)

[0012] Wherein, r is the micro-element length of the bit radius; Tf The torque consumed to overcome the wellbore friction; T w The torque transmitted to the drill bit for rock breaking and well drilling; P W The power used for rock breaking and sliding drilling during well drilling; ω is the instrument rotation speed, D is the drill bit diameter; μ is the friction factor of the drill bit; d() is the differential.

[0013] Step 2: Set the constraint conditions of the bit wear amount, the weight on bit of the upper part of the drill string, and the instrument rotation speed with the unit well drilling length cost, mechanical specific energy, and drilling measurement speed as the objective functions.

[0014] As a preferred embodiment of the present invention, the formula of the objective function is:

[0015]

[0016] Wherein, C is the unit well drilling length cost; MSE is the mechanical specific energy, v is the drilling measurement speed, and h is the bit wear amount.

[0017] As a preferred embodiment of the present invention, the value range of the bit wear amount and the maximum and minimum value ranges of the weight on bit of the upper part of the drill string and the instrument rotation speed are set in the constraint conditions.

[0018] Step 3: Optimize the objective function using the MGO algorithm.

[0019] As a preferred embodiment of the present invention, the MGO algorithm includes:

[0020] Step 31: Initialize the positions of the moss individuals.

[0021] Step 32: Determine the wind direction.

[0022] Step 33: Search for spore dispersal.

[0023] Step 34: Conduct double reproduction search.

[0024] Step 35: Conduct the cryptobiotic mechanism.

[0025] As a preferred embodiment of the present invention, the MGO algorithm is improved to obtain the IMGO algorithm, including: using the Tent chaotic map to replace the random function in the positions of the moss individuals, and the formula of the Tent chaotic map is:

[0026]

[0027] Wherein, a is the control parameter; x n is the value of the nth iteration.

[0028] As a preferred embodiment of the present invention, the IMGO algorithm further includes: introducing dynamic weights into the i-th new individual of the double reproduction search of the MGO algorithm and the corresponding j-th particle, and the formula is:

[0029]

[0030]

[0031] step3 = 0.1E(r6 - 0.5)

[0032] The formula for the dynamic weight ω is:

[0033]

[0034] where T represents the number of iterations and δ represents a constant; represents the i-th new individual; represents MO i new the j-th particle in; MO best,j represents MO best the j-th particle of; D_wind j represents the j-th particle of D-wind; r4, r5, and r6 represent random numbers.

[0035] As a preferred embodiment of the present invention, it further includes: inputting the WOB on the upper part of the drill string, the instrument rotation speed, and the bit wear amount predicted by the MGO algorithm or the IMGO algorithm into the motion model to obtain the matrix of the drilling parameters of the instrument physical entity itself, constructing the influence coefficient between the drilling parameters of the instrument physical entity itself and the drilling parameters of the digital twin model, and mapping the matrix of the drilling parameters of the instrument physical entity itself and the matrix of the influence coefficients of the drilling parameters to obtain the matrix of the drilling parameters of the digital twin model.

[0036] As a preferred embodiment of the present invention, a drilling parameter prediction system for the digital twin model of a logging-while-drilling tool based on MGO includes: a memory for storing instructions executable by a processor; and a processor for executing the instructions to implement the method for predicting the drilling parameters of the digital twin model of the logging-while-drilling tool based on MGO.

[0037] Advantages of the present invention:

[0038] 1. The present invention constructs an integrated "multi-phase multi-field digital twin model" that can reflect the real-time operating health status of the instrument during the drilling process and the influence of the stress coupling cascade of the wellbore environment on the logging response of the instrument, and can reflect the changes in multiple physical fields such as the temperature field, pressure field, and stress field in real time, as well as the distribution and evolution of the complex geological structure of the reservoir. Relying on this digital twin model, it can provide an explanatory basis for revealing the degradation of instrument sensitive parameters and the correlation mechanism of fault generation under the condition of multi-physical field coupling, as well as the evolution law of the abnormal characteristics of the logging response output;

[0039] 2. The digital twin model can simulate the evolution process of sensitive parameter degradation and early fault generation of key components of the while-drilling azimuth electromagnetic wave logging instrument, so as to give a reasonable explanation when the while-drilling geosteering signal suddenly becomes abnormal. The simulation results can provide effective technical support for real-time wellbore trajectory optimization;

[0040] 3. Construct a motion modeling method for drilling parameters of the azimuth electromagnetic wave logging instrument while drilling; perform force analysis on the drilling motion model of the instrument to obtain the mathematical model of drilling parameters and the relationship between the various parameters; the drilling motion model needs to select reasonable parameters and coefficients, which can be divided into formation coefficient, drilling parameters and drill bit structure coefficient; among them, the drill bit structure coefficient can be determined by consulting the manual according to the actual situation on site, and the formation coefficient is difficult to obtain directly, so the mathematical model of the three parameters of drilling speed, drilling length and drilling time is obtained by modifying the Young drilling speed model and the drill bit wear model, and the expression of the formation coefficient can be solved jointly, and finally the formation coefficient can be obtained by substituting the actual drilling data into the expression, and the motion model of the drilling parameters can be established;

[0041] 4. Multi-parameter and multi-objective optimization modeling method for drilling while drilling azimuth electromagnetic wave logging instrument based on IMGO; on the basis of the mathematical model of drilling parameters, the multi-parameter and multi-objective optimization model is established with drilling speed, unit drilling length cost and mechanical specific energy as objective functions, and drilling pressure, rotation speed and drill bit wear as decision variables; on the basis of MGO, the improved IMGO is obtained by adding Tent chaos map initialization and dynamic weight update strategy. The improved algorithm is verified by the test function, which is better than other previous algorithms in convergence speed, convergence accuracy and optimization ability, and can be used for multi-objective optimization more efficiently; the Pareto optimal solution set of the three objective functions is calculated by the IMGO algorithm, which provides a variety of options for multi-objective drilling parameter optimization at the drilling site;

[0042] 5. Modeling method for real-time driving of digital twin model by azimuth electromagnetic wave logging while drilling instrument; the drilling parameters optimized by IMGO algorithm constitute the main motion parameter set of azimuth electromagnetic wave logging while drilling instrument, and then the relationship with the digital twin model of the instrument is established through mapping matrix, and finally the digital twin model of multi-parameter and multi-objective optimization of azimuth electromagnetic wave logging while drilling instrument is established. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is a flow chart of the drilling parameter prediction method of the digital twin model of the logging while drilling tool based on MGO of the present invention;

[0044] Figure 2 It is a schematic diagram of the drilling motion model of the azimuth electromagnetic wave logging while drilling instrument;

[0045] Figure 3 Schematic diagram of the comparison between the torque applied to the top of the drill string and the measured value when considering and not considering the torque consumed by friction in the present invention;

[0046] Figure 4 Comparison diagram of the influence of not considering friction torque and considering friction torque on the drilling speed in the present invention;

[0047] Figure 5 Comparison diagram of the influence of not considering friction torque and considering friction torque on the cost per unit drilling length in the present invention;

[0048] Figure 6 Comparison diagram of the influence of not considering friction torque and considering friction torque on mechanical specific energy in the present invention;

[0049] Figure 7 Comparison diagram of the predicted results of the drilling speed in the present invention;

[0050] Figure 8 Comparison diagram of the predicted results of the drilling time in the present invention;

[0051] Figure 9 Pareto optimal solution set distribution of the drilling speed and the cost per unit drilling length in the present invention;

[0052] Figure 10 Pareto optimal solution set distribution of the drilling speed and the mechanical specific energy in the present invention;

[0053] Figure 11 Pareto optimal solution set distribution of the mechanical specific energy and the cost per unit drilling length in the present invention;

[0054] Figure 12 Convergence curve diagram of different algorithms under different test functions. Detailed implementation manners

[0055] The present invention will be further described below in conjunction with the accompanying drawings and embodiments. This figure is a simplified schematic diagram, which only illustrates the basic structure of the present invention in a schematic manner, so it only shows the components related to the present invention.

[0056] As Figure 1 shown, the drilling parameter prediction method based on the digital twin model of the logging-while-drilling instrument based on MGO includes the following steps:

[0057] Step 1: Construct a motion model of the azimuthal electromagnetic wave logging-while-drilling instrument to obtain drilling parameters;

[0058] The drilling parameters include: drilling speed v, instrument rotation speed ω, power P of the turbine mud generator, pitch angle θ, weight on bit F1 at the upper part of the drill string, and instrument torque T;

[0059] AsFigure 2 During the drilling process, the drill string is simplified as a homogeneous and isotropic material. The external guide tube (the gray arc part in the figure) mainly transmits the drilling pressure, and the azimuthal electromagnetic logging-while-drilling tool and the drill string are used for directional sliding drilling.

[0060] Taking the center of curvature of the bent drill string as the origin O, the radial coordinate axis as the R-axis, the normal coordinate axis as the T-axis, and the axial coordinate axis as the Z-axis, a cylindrical coordinate system is established. Among them, r is the radius of curvature of the wellbore, m; F1 represents the drilling pressure at the top of the drill string, N; F2 represents the drilling pressure at the bottom of the drill string, kN; T is the torque applied at the top of the drill string, N·m; M represents the elemental gravity of the drill string, kN; F n and f represent the supporting force and frictional force of the drill string against the wellbore during drilling, kN; θ represents the angle between each elemental section and the normal axis, °.

[0061] The azimuthal electromagnetic logging-while-drilling tool drills at a constant speed against the frictional force of the wellbore wall. The drilling path is differentiated into dθ, the force analysis is carried out on each elemental section, and then the integration of the drilling path from 0 to π / 2 is performed to obtain the force state of the entire path on the radius of curvature of the wellbore. The formulas for the drilling pressure F2 at the bottom of the drill string and the frictional force f are:

[0062]

[0063] f = F1 - F2 (2)

[0064] Among them, ρ is the density of the external material of the logging-while-drilling tool, kg / m; g is the acceleration due to gravity, m / s 2 ; μ is the friction coefficient of the drill bit.

[0065] From the modified Young's drilling rate model, the relationship between the drilling speed v and the drilling pressure F1 at the top of the drill string and the rotational speed ω can be obtained:

[0066]

[0067] Among them, K f is the formation drillability coefficient; C p is the bottom hole pressure differential influence coefficient; C H is the hydraulic purification coefficient; F M is the threshold drilling pressure, kN; ω is the rotational speed, r / min; C2 is the drill bit wear factor; h is the drill bit wear amount.

[0068] During the drilling process, ignoring the influence of the drilling fluid column pressure on drilling, the power P of the turbo mud generator is constant, and the power used to overcome the frictional force of the wellbore wall is denoted as P f , and the power used for rock breaking and sliding drilling during the drilling process is denoted as P W The formula for is:

[0069] P f = fv (4)

[0070] P W = P - P f (5)

[0071] The torque applied to the top of the drill string, after overcoming the wellbore friction, drill string buckling, and inertial forces, transfers the torque to the drill bit to break the rock and drill the well. Without considering the influence of drill string buckling and inertial forces, a drill string torque model is established. The formula for the torque T applied to the top of the drill string is as follows:

[0072]

[0073]

[0074] T = T f + T w (8)

[0075] where r is the micro-length of the drill bit radius, in mm; T f is the torque consumed to overcome the wellbore friction, and T w is the torque transmitted to the drill bit to break the rock and drill the well; D is the drill bit diameter; d() is the differential.

[0076] In existing drilling engineering, only T w is considered, and T f is not considered. However, the sliding friction force of the drill bit is an influential factor that cannot be ignored in torque calculation. Therefore, the present invention uses the sliding friction coefficient μ and the weight on bit F1 at the top of the drill string to calculate the torque T f consumed to overcome the wellbore friction.

[0077] As Figures 3 - 6 shown, Figure 3 it can be seen that the torque T applied to the top of the drill string is closer to the measured torque after considering T f . Without considering the friction torque, it has a greater impact on the optimization result of the final objective function, as Figures 4 - 6 shown; therefore, the torque consumed to overcome the wellbore friction needs to be considered.

[0078] By studying the various factors affecting the drill bit wear amount h, the drill bit wear model is obtained as follows:

[0079]

[0080] where A f is the formation abrasiveness coefficient; D1 and D2 are the weight on bit influence coefficients; C1 is the drill bit wear slowdown coefficient, and a1 and a2 are the drill bit type coefficients.

[0081] Dividing Equation (3) by Equation (9) gives:

[0082]

[0083] Integrating both sides of Equation (10) gives the relationship between the drilling length y and the bit wear h:

[0084]

[0085] where y is the drilling length, in m.

[0086] Integrating both sides of Equation (9) gives the relationship between the drilling time t and the bit wear h:

[0087]

[0088] where t is the drilling time, in h.

[0089] In the application scenario, the bit of the azimuthal electromagnetic logging-while-drilling tool uses insert bits, so C1 = C2 = 2, and then Equations (11) and (12) can be simplified to:

[0090]

[0091] From Equation (13), it can be solved that:

[0092]

[0093] In the actual drilling scenario, the bit type at the front of the azimuthal electromagnetic logging-while-drilling tool is HJ517, and the detailed parameters are shown in Table 1.

[0094] Table 1 Bit-related coefficients

[0095]

[0096] In Table 1, D1 and D2 are bit size coefficients, a1 and a2 are bit type coefficients, C1 is the bit wear slowdown coefficient, and C2 is the bit wear factor.

[0097] Assume that there is no bottom-hole pressure difference and the bottom-hole purification is sufficient, then the bottom-hole pressure difference influence coefficient C p = 1, and the hydraulic purification coefficient C H = 1; in the motion model, in addition to the above-known coefficients and parameters, there are also the formation abrasiveness coefficient A f and the formation drillability coefficient K f which have not been calculated yet. During the drilling process, these coefficients will change with the formation. To make the model closer to the actual situation, a section of drilling-related data from 2513 to 2646 m in a certain area is selected, as shown in Table 2. In this section of data, the formation parameters have no obvious changes and can be used to obtain the two unknown parameters.

[0098] Table 2 Drilling-related data in a certain area

[0099]

[0100] Table 2 shows the drilling - related data for a section from 2513 m to 2646 m in a certain area. The data includes the drilling length y, the drilling time t, the drilling speed v, the weight on bit F1, and the rotational speed ω.

[0101] Calculate the formation abrasiveness coefficient and the formation drillability coefficient for the section from 2513 m to 2646 m; in order to maximize the utilization rate of the drill bit and ensure the safety and reliability of the drilling process, the drill - bit wear is set at 0.7.

[0102] Calculating the average weight on bit and the average rotational speed and substituting them into Equation (14) gives

[0103] Step 2: Based on the established motion model, during the drilling process, not only need to consider how to complete the drilling task as soon as possible, but also need to minimize the drilling cost and energy consumption as much as possible, that is, hope to have the minimum cost per unit drilling length, the minimum mechanical specific energy MSE, and the maximum drilling speed; find the best balance among the three decision variables of the weight on bit, the drilling speed, and the drill - bit wear, and establish a multi - objective drilling parameter optimization model with three objective functions and constraint conditions;

[0104] The three objective functions are:

[0105] After standardizing the objective functions, we get:

[0106] For the overall solution to be practical and feasible, the three decision variables should be within the feasible range. So the constraint conditions for the three decision variables are as follows:

[0107] 1. The constraint condition for the drill - bit wear h is: 0 ≤ h ≤ 1;

[0108] 2. The constraint condition for the weight on bit F1 is: F 1min ≤ F1 ≤ F 1max , F 1min =50 kN, F 1max =150 kN;

[0109] 3. The constraint condition for the rotational speed ω is: ω min ≤ ω ≤ ω max , ω min =50 r / min, ω max =100 r / min;

[0110] During the drilling process, it is crucial to consider the cost per unit drilling length. The cost is divided into the drill - bit cost and the rig cost, which are expressed by the following expressions:

[0111]

[0112] Where C is the cost per unit drilling length, in yuan / m; C b is the bit cost; t is the working time, in h; y is the drilling length, in m.

[0113] Substituting Equation (13) into Equation (15) gives:

[0114]

[0115] The mechanical specific energy MSE is the mechanical energy required for the bit to break a unit volume of rock per unit time. In actual engineering, the mechanical specific energy is closely related to the rock-breaking efficiency and is an important indicator for engineers to evaluate the drilling efficiency; the expression is:

[0116]

[0117] Where MSE is the mechanical specific energy, in MPa; ω is the rotational speed, in r / min; v is the drilling speed, in m / h; F1 is the weight on bit, in kN.

[0118] The drilling and logging speed is as shown in formula (3).

[0119] Step 3. Construct a Moss Growth Optimization (MGO) algorithm to optimize the objective function;

[0120] Moss usually grows in dark and humid places. It has three reproduction methods: asexual reproduction, sexual reproduction, and vegetative reproduction; the unique survival mechanism of moss: the cryptobiotic mechanism, that is, when moss faces drought or other threats, it will enter a dormant state, and when the conditions are favorable, the moss will continue to grow.

[0121] Step 31. Initialize the position of the moss individual:

[0122]

[0123] Where MO i,j represents the position of the i-th moss individual in the j-th dimension, u b and l b represent the upper and lower bounds, N represents the scale, and D’ represents the dimension;

[0124] Preferably, the Tent chaotic mapping is used to initialize the position of the moss individual;

[0125]

[0126] Among them, a is the control parameter, and different control numbers are determined according to the actual situation, representing different chaotic strategies; in this embodiment, a = 0.6; x n is the value of the n-th iteration, ranging between [0, 1]; that is, using T(xn ) Replace the rand(0,1) function.

[0127] Step 32: Determine the wind direction;

[0128] The wind spreads the spores of moss everywhere, playing an important role in the reproduction of moss. In the MGO algorithm, the wind direction determination mechanism makes the wind direction always blow from the location where most mosses are to the location most favorable for growth, thus preventing the MGO algorithm from falling into a local optimal solution. The precise calculation of the wind is as follows:

[0129]

[0130] In the formula, D-wind represents the wind direction; num represents the total number of individuals in dirX, and dirX represents the set of distances of individuals in dirX relative to MO i Relative to MO best of the distance set.

[0131] Step 33: Spore dispersion search;

[0132] Under the spread of the wind, the spread of spores has great uncertainty. Part of it spreads under the condition of stable wind, and the other part spreads under the condition of turbulence. This enables the MGO algorithm to allow individuals to make random choices, increasing the types of step sizes and accelerating the convergence speed of the algorithm, ultimately forming the diversity of the population. The mathematical model is as follows:

[0133]

[0134] In the formula, represents a new moss obtained by spreading spores from the i-th moss individual MO i ; r1, r2, and r3 represent random numbers within (0,1); E represents the wind intensity. When r1 > 0.2, the spores disperse under stable wind and the step size step1 is selected; when r1 ≤ 0.2, the spores disperse under turbulent wind and the step size step2 is selected;

[0135] Step 34: Dual reproduction search;

[0136] After the spores complete their spread, the moss undergoes sexual reproduction and vegetative reproduction to produce new individuals. Different from traditional algorithms that search in a single dimension, the MGO algorithm searches in two dimensions, enhancing the local search efficiency of the algorithm. The specific mathematical model is as follows:

[0137]

[0138] In the formula, represents the i-th new individual, represents the j-th particle in best,j represents MObest the j-th particle of, D_wind j represents the j-th particle of D-wind; r4, r5, and r6 represent random numbers within (0, 1); when r4 > 0.5, sexual reproduction is used for search; when r4 ≤ 0.5, vegetative reproduction is used for search;

[0139] After initializing the population, the moss spores in the population are spread according to the wind direction. The subsequent double reproduction search may search near the global optimum point, resulting in the premature convergence of the algorithm. To reduce the probability of this phenomenon occurring, a dynamic weight ω is added in the double reproduction search stage. In the initial stage of optimization, a larger dynamic weight ω is given when the number of iterations is small to ensure the global search ability. In the later stage of optimization, a smaller dynamic weight ω is given when the number of iterations is large to ensure the local search ability and improve the convergence speed;

[0140] The formula for the moss double reproduction search after optimizing the dynamic weight is:

[0141]

[0142] The formula for the dynamic weight ω is:

[0143]

[0144] In the formula, T represents the number of iterations, T max is the maximum number of iterations, and δ represents a very small constant used to avoid the weight ω = 0.

[0145] Step 35, cryptobiosis mechanism;

[0146] The cryptobiosis mechanism refers to the phenomenon of moss reviving after drought for a certain period of time. Different from the method of directly modifying individuals in traditional meta-heuristic algorithms, the cryptobiosis mechanism simulates the growth of moss. This mechanism will record the individuals in each iteration. After completing the population iteration, the optimal individual will be selected to replace the current individual, thus avoiding falling into the local optimal solution.

[0147] Based on the moss growth optimization algorithm, the present invention adds Tent chaos mapping and dynamic weight update strategy to ensure the convergence speed and search accuracy of the algorithm, and obtains the improved moss growth optimization algorithm (IMGO).

[0148] It further includes: Step Four, real-time driving of the digital twin model;

[0149] Input the hook load F1, instrument rotation speed ω, and bit wear amount predicted by the IMGO or MGO algorithm into the motion model in Step One to construct the drilling parameter matrix A1. The formula is:

[0150] A1 = [v ω P θ F1 T] (25)

[0151] Construct a functional mapping relationship between drilling parameters and instrument entities, and the mapping matrix of the mapping relationship is expressed as:

[0152]

[0153] Among them, K 11 represents the influence coefficient of the drilling and logging speed v of the instrument physical entity itself on the drilling and logging speed v' in the digital twin model; K 12 represents the influence coefficient of the drilling and logging speed v of the instrument physical entity itself on the rotational speed ω' in the digital twin model; K 13 represents the influence coefficient of the drilling and logging speed v of the instrument physical entity itself on the output power P' of the generator in the digital twin model; K 14 represents the influence coefficient of the drilling and logging speed v of the instrument physical entity itself on the pitch angle θ' in the digital twin model; K 15 represents the influence coefficient of the drilling and logging speed v of the instrument physical entity itself on the drill pressure F1' of the upper part of the drill string in the digital twin model; K 16 represents the influence coefficient of the drilling and logging speed v of the instrument physical entity itself on the instrument torque T' in the digital twin model; similarly, K 21 to K 26 represent the influence coefficients of the rotational speed ω of the instrument physical entity on v', ω', P', θ', F1', and T' in the digital twin model in turn; K 31 to K 36 represent the influence coefficients of the output power P of the instrument physical motor on v', ω', P', θ', F1', and T' in the digital twin model in turn.

[0154] Through the mapping relationship, the drilling parameter matrix of the digital twin model of the azimuth electromagnetic logging-while-drilling instrument can be obtained:

[0155] A1' = K1A1 (27)

[0156] Similarly, mapping relationships can also be established for the motion modes of the digital twin models of other components (such as transceiver coils, circuit boards, generators, etc.), and their mapping matrices can be represented by K i Then, for the digital twin model motion mapping matrix K of the n components of the entire azimuth electromagnetic logging-while-drilling instrument, it can be expressed as:

[0157]

[0158] For the n components of the entire instrument, the motion parameters of their physical entities can be expressed as:

[0159] A = [A1 A2…A n (29)

[0160] Then, the motion parameters of the n components of the instrument mapped to the digital twin model can be expressed as:

[0161] A' = KA (30)

[0162] So far, a digital twin that can map the operating state of the azimuth electromagnetic logging-while-drilling instrument in real time has been established, realizing the real-time drive of the physical entity of the instrument to the digital twin model.

[0163] Verify the accuracy of the model:

[0164] Substituting the obtained formation abrasiveness coefficient and formation drillability coefficient into equations (3) and (13), the predicted drilling speed and drilling time in the motion model can be obtained, as Figure 7 and Figure 8 shown.

[0165] From Figure 7 and Figure 8 it can be seen that the root mean square error (RMSE) and mean absolute error (MAE) of the drilling speed and drilling time are within the allowable range of the sample error, and the predicted data conform to the drilling law, so the subsequent drilling parameter optimization can be carried out; Figure 7 represents the comparison between the predicted value and the true value of the drilling speed v obtained from the motion model, and the root mean square error (RMSE) and mean absolute error (MAE) are 0.501 and 0.396 respectively; Figure 8 represents the comparison between the predicted value and the true value of the drilling time t obtained from the motion model, and the root mean square error (RMSE) and mean absolute error (MAE) are 0.597 and 0.475 respectively.

[0166] Figure 9 、 Figure 10 and Figure 11 are the distribution of the Pareto optimal solution sets of the three objective functions of drilling speed, unit drilling length, and mechanical specific energy under IMGO; among them, the blue points are the Pareto optimal solution sets of IMGO, and the red points are the solution sets before optimization. IMGO finds more optimal solutions and the distribution is more uniform than before optimization;

[0167] Randomly select 5 groups of solution sets from the Pareto optimal solution sets of IMGO, and list the corresponding numerical values of relevant parameters, as shown in Table 3, where the first 5 groups of data are the values after IMGO optimization, and the 6th group of data is the value before optimization.

[0168] Table 3 Optimized parameter values of randomly selected 5 groups of Pareto optimal solution sets

[0169]

[0170] Table 3 shows that 5 groups of data randomly selected from the Pareto optimal solution set distribution of the three objective functions are the optimized data, numbered 1-5 respectively, and number 6 is the data before optimization.

[0171] The decision variable values are the weight on bit F1 (kN), the rotational speed ω (r / min), and the bit wear h.

[0172] After comparison, it can be seen that IMGO has achieved obvious effects in accelerating the drilling speed, reducing the cost per unit drilling length, and reducing energy consumption; among them, the drilling speed has increased by 19.71%, the cost per unit drilling length has decreased by 22.48%, and the mechanical specific energy has decreased by 23.21%.

[0173] Figure 12 The fitness values of the Sparrow Search Algorithm (SSA), Grey Wolf Optimization Algorithm (GWO), Moss Growth Optimization Algorithm (MGO), and Improved Moss Growth Optimization Algorithm (IMGO) under 4 different functions are used to test the performance of the algorithms under different functions; among them, Figure 12 f1-f2 in a and b are unimodal functions, Figure 12 f5-f6 in c and d are multimodal functions, aiming to test the performance of the algorithms under different types of test functions; to ensure fairness in comparison, the population size of the algorithms is set to 30, the dimension is 30, the maximum number of iterations is 1000, and the convergence curves are obtained after each algorithm operates independently 30 times.

[0174] From Figure 12 the convergence curves, it can be seen that the IMGO algorithm has better performance than the SSA, GWO, and MGO algorithms in terms of optimization ability, convergence speed, and convergence accuracy; it can be used for the multi-objective optimization problem of drilling parameters.

[0175] Taking the ideal embodiments of the present invention described above as an inspiration, through the above description, relevant staff can completely make various changes and modifications without departing from the technical idea of this invention. The technical scope of this invention is not limited to the content in the specification, and its technical scope must be determined according to the scope of the claims.

Claims

1. A drilling parameter prediction method based on the digital twin model of the logging while drilling tool based on MGO, characterized in that: The following steps are involved: Step 1: construct a motion model of the while-drilling azimuth electromagnetic wave logging instrument to obtain drilling parameters; Step 2: Taking the unit drilling length cost, mechanical specific energy and drilling speed as the objective function, set the constraints of drill bit wear, drilling pressure on the upper part of the drilling tool and instrument speed; Step 3: Use the MGO algorithm to optimize the objective function.

2. The method for predicting drilling parameters of a digital twin model of a logging while drilling tool based on MGO according to claim 1, characterized in that: The drilling parameters include drilling speed v, instrument rotation speed ω, turbine mud generator power P, pitch angle θ, drilling pressure F1 on the upper part of the drilling tool, and instrument torque T.

3. The method for predicting drilling parameters of a digital twin model of a logging while drilling tool based on MGO according to claim 2, characterized in that: The formula for instrument torque T is: T=T f +T w (8) Where r is the microelement length of the drill bit radius; T f The torque consumed to overcome the friction of the well wall; T w P is the torque transmitted to the drill bit for breaking rocks and drilling; W is the power used for breaking rocks and sliding drilling during drilling; ω is the instrument speed, D is the drill bit diameter; μ is the friction factor of the drill bit; d() is the differential.

4. The method for predicting drilling parameters of a digital twin model of a logging while drilling tool based on MGO according to claim 1, characterized in that: The MGO algorithm includes: Step 31, initializing the position of individual mosses; Step 32, determine the wind direction; Step 33, searching for spore dispersal; Step 34, perform double breeding search; Step 35: Perform cryptobiotic mechanism.

5. The method for predicting drilling parameters of a digital twin model of a logging while drilling tool based on MGO according to claim 4, characterized in that: The MGO algorithm is improved to obtain the IMGO algorithm, including: using the Tent chaotic map to replace the random function in the moss individual position. The formula of the Tent chaotic map is: Among them, a is the control parameter; x n is the value of the nth iteration.

6. The method for predicting drilling parameters of a digital twin model of a logging while drilling tool based on MGO according to claim 1, characterized in that: The IMGO algorithm also includes: introducing dynamic weights in the i-th new individual and the corresponding j-th particle in the double reproduction search of the MGO algorithm, the formula is: step3=0.1E(r6-0.5) The dynamic weight ω formula is: Where T represents the number of iterations and δ represents a constant; represents the i-th new individual; express The jth particle in MO best,j means MO best The jth particle of j represents the j-th particle of D-wind; r4, r5 and r6 represent random numbers.

7. The method for predicting drilling parameters of a digital twin model of a logging while drilling tool based on MGO according to claim 1, characterized in that: The formula of the objective function is: Among them, C is the cost per unit drilling length; MSE is the mechanical specific energy, v is the drilling speed, and h is the drill bit wear.

8. The method for predicting drilling parameters of a digital twin model of a logging while drilling tool based on MGO according to claim 1, characterized in that: The constraint conditions include setting the range of drill bit wear and the maximum range of drilling pressure on the drill bit and instrument rotation speed.

9. The method for predicting drilling parameters of a digital twin model of a logging while drilling tool based on MGO according to claim 1, 5 or 6, characterized in that: Also includes: The upper drilling pressure of the drill tool, instrument rotation speed and drill bit wear predicted by the MGO algorithm or the IMGO algorithm are input into the motion model to obtain the drilling parameter matrix of the instrument physical entity itself, and the influence coefficients of the drilling parameters of the instrument physical entity itself and the drilling parameters of the digital twin model are constructed. The drilling parameter matrix of the instrument physical entity itself and the influence coefficient matrix of the drilling parameters are mapped to obtain the drilling parameter matrix of the digital twin model.

10. The drilling parameter prediction system of the digital twin model of the logging while drilling tool based on MGO is characterized by: include: a memory for storing instructions executable by a processor; A processor, used to execute instructions to implement the drilling parameter prediction method of the digital twin model of the MGO-based logging while drilling tool as described in any one of claims 1 to 9.

Images